Nanda - 2001 - Rice breeding and genetics research priorities an

ABOUT THE BOOK
Rice is one of the major cereals of the world and is the staple for about 2.7
billion people in Asia. Global demand for rice is projected to grow at least
commensurately with the population growth. A 70 percent increase in
supply by the year 2025 will be required to maintain the food-population
balance.
The green revolution of the Í960's and 1970'sin particular in Asia brought
about a significant Increase in rice production in the irrigated ecology. It
established without doubt the technical feasibility of maintaining rice
production well ahead of the population growth in some of the developed
and underdeveloped rice growing countries. However this phenomenal
growth in production achieved with the adoption of green revolution
technologies was not without its adverse effects; it has generated
numerous problems ranging from biological to environmental and socioeconomic.
Another area of concern is that the rice yield in the rainfed
ecologies has remained stagnant over the years. There is a need for
innovative approaches to enhance rice production across ,the rice
ecologies to meet future challenges.
This book has sixteen chapters. Scientists of international repute from
International Research Institutes, Universities and Research Foundations
have contributed the chapters in this book. The book enumerates past
achievements, the future prospects and possible approaches in the field of
'Rice Breeding and Genetics'. The book will be of immense use to research
scientists and policy makers.
ISBN 1-57808-086-X

Preface
The International Rice Research Institute and the Food and Agriculture
Organization of the United Nations have carried out critical studies and
projected the future demand of rice and outlined various approaches to
meet the challenges. I have tried in this publication to compile research
priorities in the field of . genetics and plant breeding to enhance rice
production to meet future challenges.
I am grateful to Drs. G. S. Khush, O. Ito^ S. S. Virmani^ D. Senadhira,
Gloria Cabuslay, Evangelina Ella, and R. C. ChaudJhary former global
co-ordinator, INGER of the International Rice Research Institute;
Dr. B. N. Singh from the International Institute of Tropical Agriculture;
Dr. H. Ikehashi from the University of Kyoto, Japan; Dr. M. J. Lawrence
from the University of Birmingham, UK; Drs. A. P, K. Reddy and J. S.
Bentur from the Rice Directorate, Hyderabad, India; Drs. S. D. Sharma
and J. Biswal from M. S. Swaminathan Research Foundation, Chennai,
India; Drs. S. R. Dhua and P. K. Agarwal from the Central Rice Research
Institute, Cuttack, India; Dr. K. K, Jena from Mahyco Research
Foundation, India; Dr. R. J. Singh from the University of Illinois, Urbana,
USA; and Prof. Zhang Yu from Huazhong Agricultural University,
Wuhan, China for their quick response and willingness to contribute
chapters for the book in their field of specialization. The quality of the
book is undoubtedly enriched due to their contributions. My thanks are
due to the Food and Agriculture Organization of the United Nations for
the permission accorded to reproduce two chapters namely the Key
Note Address by F.Riveros and Sustainable Integrated Rice Production
by Sastry et ah from the Proceedings of the 18th Session of the
International Rice Commission.
The inspiration to compile this publication dawned on me during a
visit to Moline, Illinois, USA, in January 1997 to welcome my grand
daughter, Ayesha, This publication is therefore dedicated to her, Ayesha,
to cherish her arrival.
I take the opportunity to extend sincere thanks to my wife,
Anupama, my critic as well as source of encouragement. My special

vi
Rice Breeding and Genetics; Research Priorities and Challenges
thanks go to our children Jeetu, Upasana^ Dolly/ Aruu/ Jolly and Praha
for their expert assistance in computer use and constant support.
4th January 1999
Jata S. Nanda
860/ Colony Lake Drive/
Schaumburg/ IL. 60194
USA

Contents
Preface
Keynote Address of the 18th Session of IRC
F. Riveros
1. Rice Breeding and Genetics: Research Perspectives
J.S. Nanda
2. Hybrid Rice
J,S. Nanda and S.S. Virmani
3. Sustainable Integrated Rice Production
S.V, Shasfty^ D.V. Tran, V,N. Nguyen and J.S. Nanda
4. Drought and Submergence in Rice Production
Osamu Ito, Gloria Cabuslay and Evangelina Ella
5. New Plant Type of Rice for Increasing the Genetic
Yield Potential
Gurdev S. Khush
6. Hybrid Sterility in Rice—^Its Genetics and Implication
to Differentiation of Cultivated Rice
H. Ikehashi
7. A Critical Evaluation of Current Breeding Strategies
M./. Lawrence and D. Senadhira
8. Insect and Disease Resistance in Rice
A.P.K. Reddy and }.S, Bentur
9. Breeding Rice for Resistance to Diseases and Insect Pests
Ram C. Chaudhary
10. Breeding for Adverse Soil Problems in Rice
B.N. Singh
11. Molecular Marker-Based Gene Tagging and
Its Impact on Rice Improvement
Qifa Zhang and Sibin Yu
12. Exploitation of Alien Species in Rice Improvement—
Opportunities, Achievements and Future Challenges
V
1
9
23
53
73
99
109
119
143
165
215
241
269

viii
Rice Breeding and Genetics: Research Priorities and Challenges
13. Cytogenetics of Rice 285
RJ. Singh and G.S, Khush
14. Species of Genus Oryza and Their Interrelationships 311
S.D. Sharma, S.R. Dhua and P.K. Agarwal
15. Origin of O. Sativa and Its Ecotypes 347
S.D. Sharma, Smita Tripathy and Jyostnamayee Biswal
Index 371

Keynote Address of the
18th Session of IRC
F. Riveros’^
Rice is the staple food of more than half of the world^s population. In
Asia alone^ 90% of the wofld's rice is produced and consumed. Most of
the consumers, who depend on rice as their primary food, live in less
developed countries. It is foreseen that the world's population may
exceed 8 billion by 2025 and will need about 765 million tons of rice, 70%
more than what is consumed today. This increase in rice production
must be achieved through utilization of less land, less water, fewer
agrochemical and other inputs. It is thus imperative to find ways and
means to lift the present yield level, optimize the use of various inputs
such as water and fertilizer in order to make the rice production efficient,
cost effective, suitable for resource-poor farmers, sustainable and
environment friendly.
The green revolution of the 1960s and 1970s brought significant
increases in rice productivity and rice production. In countries where
rice is a staple food, production has increased by an average of 70% over
the last 25 years. This has been possible through irrigation, the
popularization of agrochemicals (fertilizers and pesticides), and genetic
improvement of the rice plant to enhance its yield potential and
tolerance to biotic and abiotic stress. The merits of semidwarf highyielding
varieties (HYVs) such as high-tillering, non-lodging habit,
better interception of solar radiation and improved harvest index have
contributed to the wide acceptance of the HYVs by farmers in the
tropics. The extensive adoption of HYVs and improved production
technology were accelerated through favourable government policies;
expansion of irrigated area; availability of credit facilities; intensive

. XVjiSSSS?.
2 Rice Breeding and Genetics: Research Priorities and Challenges
extension services; and availability of agrochemicals, especially
fertilizer. During this period, the seed industry gained: momentum in
many countries, with increased availability of improved quality seed.
The research infrastructures were also strengthened. With time, the
intrinsic management-responsive merit of HYVs was recognized.
Cropping intensity was also increased due to the introduction of earlymaturing
photoperiod-inserisitive varieties. However, the wide
adoption of HYVs in place of traditional rice is the main cause of genetic
erosion and the decline in natural biodiversity.
The green revolution undoubtedly established the technical feasibility
of maintaining rice production well ahead of the population growth
in many developing countries. Nevertheless, it virtually bypassed the
countries in sub-Saharan Africa, made limited contributions in countries
with poor water control and totally failed in areas with problem soils.
Recent observations, however, have shown a fall in gains and signs of
stress in intensely cultivated irrigated lands. In order to meet the everincreasing
demand, about 10 million tons of more rice per year has to be
produced. New technology frontiers need to be explored to increase rice
production in support of food security, especially in low-income fooddeficit
countries (LIFDCs). Environment-friendly socioeconomically
acceptable technologies need to be developed to optimize the efficient
use of water, fertilizers and other inputs, and to ervhance productivity.
Rice is grown widely in four ecosystems: irrigated land, rainfed
lowland, upland, and deepwater and tidal swamps. Many factors
determine the relative contribution of each ecosystem to future rice
supplies; for example, production potential and chances for its
enhancement; public and private investment in production and
infrastructure; availability of water, availability and price of production
inputs; rate at which irrigated and rainfed lowland areas can be
expanded; actual and predicted prices of rice and alternative crops; and,
of course, the population growth. Irrigated rice constitutes about half
the total harvested area but contributes more than two-thirds of the total
production. Significant yield increases are expected to be realized in the
years to come from the use of the new plant type developed by the
International Rice Research Institute (IRRI), coupled with improved
resource management. The increased production resulting frorn changes
in plant type and cropping efficiency adds very little to farmer's costs.
Producers and consumers can share gains. The new plant type
developed by IRRI has a high harvest index, which was achieved by
increasing panicle size (i.e., more and bigger grains) and reducing the
number of tillers per hill.

F. Riveros 3
Hybrid rice technology offers yet another means of increasing
irrigated rice potential. China has been successful in exploiting hybrid
rice potential. Hybrid rice yields on average 15-20% more than the
common improved rice varieties. Recent rice hybrids have increased
yields 30-40% (about 2t ha'^) beyond the limits set by the improved
semidwarf varieties with the use of genetically diverse parental lines;
japónica x indica or javanica x indica crosses are especially promising,
China and IRRI have capitalized on this technology. China expanded the
area planted to hybrid rice from 9.6 million ha (30%) in 1986 to around
18 million ha (or 54%) in 1993, and has saved more than 2 million ha of
land for agricultural diversification. Hybrid rice technology has not,
however, spread to other countries, which is partly due to the high cost
of hybrid seed. Practices to reduce the cost of hybrid seed and to
increase seed yield have been developed and are now being tried in a
number of countries. Exploitation of hybrid rice technology in irrigated
ecosystems in India and Vietnam appears promising in the immediate
future. There is need to improve the grain quality and incorporate a high
level of resistance to insect pests and diseases in hybrid rice to make it
widely acceptable in tropical conditions. Use of the two-line method of
hybrid seed, using temperature-sensitive male sterility (TGMS) and
photoperiod-sensitive genetic male sterility (PGMS) may further
expedite the spread of hybrid rice technology to other countries. Biotechnology
may contribute significantly to enhancing rice production though
incorporation of the apomictic trait in rice, which will allow farmers to
save their own hybrid seed.
Because of intensive cropping, especially in the irrigated lowlands
of Asia, growth in rice yield has levelled off and, in some cases, declined.
This trend needs to be reversed. In addition to the urgency of breaking
the yield barriers, the productivity of other inputs such as water and
fertilizer must also be increased. With the current irrigated rice
technology, only about 30-50% of the nitrogen fertilizer applied are
actually used by the rice plant.
In irrigated systems, more than 5000 litres of water are used to
produce 1 kg of rice. There is clearly a need to enhance the input/output
efficiency of fertilizer, water and labour with an emphasis on low-cost,
low-input, high-productivity technology. Water is a critical resource in
rice culture. The costs of bringing new areas under irrigation and
rehabilitating the existing systems are high. It is, thus important to raise
water-use efficiency in rice production systems through appropriate
water control and management techniques.
Prices of fertilizers have soared. The fertilizer industry is based on
nnn-rpni»wahlp ffos.sil fuell enerev sources and serious questions are

4 Rice Breeding and Genetics: Research Priorities and Challenges
growth that measures up to current and future needs. The promotion of
fertilizer also causes misgivingSy although no country can afford to do
away with them. Substitutes and improvement of their efficiency would
be beneficial. Agronomic practices that synergize with management of
soily water, symbiont and crop plants will have to be employed.
Integrated nutrient management systems (INMS), which consist of
improved crop agronomy to increase and sustain crop production, need
to be empahsized. In the irrigated ecology risk of salinization and soil
degradation is increasing which needs to be addressed expeditiously.
Pesticides are significant inputs to rice production. In 1988 alone, US
$ 910 million worth of insecticides were used in rice worldwide.
Integrated pest management (IPM) drastically reduces investment in
insecticides. It is environment-friendly and the risk of health hazards is
reduced. Hence adoption of IPM practices should be encouraged.
Similarly, farmers worldwide spend US $900 million each year on
herbicides to control weeds. Perhaps the expenditure on herbicides
could be decreased through the incorporation of plant traits such as
competitiveness of rice against weeds, particularly during early crop
establishment of new varieties. Exploring the use of natural pathogens
for biological control, allelopathy and proper tillage practices needs
greater emphasis to control weeds across all the ecosystems.
In addition, biotechnology has the potential to provide breeding
speed and efficiency, genetic specificity and genetic novelty to rice,
relating to productivity constraints such as diseases, insect pests, abiotic
stresses and post-harvest quality. The greatest promise in rice
improvement stems from DNA-marker-based gene identification and
insertion to create transgenic rice. The advent of molecular markers is
expected to modify rice breeding drastically, since they enable the
number of basic progenitors to be narrowed down and allow the
creation of rice genetic maps.
Collaborative research efforts in this new area should be encouraged
to develop rice that is resistant to and tolerant of major intractable
biotic and abiotic problems.
Mechanization is relevant in crop intensification and labour
enhancement and is an important component of technological development
for boosting rice production in the years to come. Equipment
needs to be developed to meet the requirements of small farmers, in
particular women farmers, to lessen the drudgery of farm operations. It
should not only be cheap but also fabricated locally so as to generate
rural employment. Technologies or more information on ''value
addition' at the farm or village level is needed to enhance incomes and

F. Riveros 5
industry: small-scale rural-based equipment makers and artisans. They
are flexible^ quick to adopt useful technologies, and provide local repair
and backup services to farmers. The necessity of decentralizing the
manufacturing industry so as to reach the rural community needs to be
stressed, as this would generate rural employment and income.
Among the various agricultural production systems, the rice-wheat
cropping systems in Asia are important both agroecologically and
socioeconomically.' This is the most extensive cropping system in the
whole world and is practised on about 23 million ha; almost 10 million
ha are in East Asia and 13 million ha in South Asia. It supports a large
number of subsistence farmers and has shown remarkable resilience in
productivity growth during the last three decades. However, there is
now a growing concern about the sustainability of the rice-wheat
production system as growth rate is stagnating and there is a tendency
towards drop in productivity. There are also problems of degradation of
soil structure; late planting of wheat, late transplanting of rice,
micronutrient deficiencies, and increased diseases and pests, especially
weeds. Accordingly, it is felt that rice-and wheat-growing countries
must accord priority to raising and sustaining the productivity of the
region^s most important production system.
The concept ^thriving with rice' seeks to increase rice production
and yield in irrigated areas. This does not solely address the rice crop,
but the whole production and processing system, involving other
agricultural activities relating to rice such as rice-cum-fish and animal
husbandry. This concept is a four-pronged approach which includes: (i)
improvemerit of economic returns of existing rice production schemes;
(ii) diversification and intensification of rice-based farming systems; (iii)
utilisation of the whole plant biomass of rice; and (iv) transformation of
rice and its by-products into value-added products. This approach was
recently introduced in irrigated rice areas in Burkina Faso, Guinea, Mali
and Senegal, In these intervention areas the rice yield was significantly
raised and rural employment was generated through development of
small farm equipment and non-agricultural services. This concept needs
to be popularized in areas where the effects of the erstwhile green
revolution are not realized.
The less favourable environments such as unfavourable rainfed
lowlands, uplands and deepwater and tidal wetlands produce 20-25%
of the world's rice. In the next two decades, they must sustain many
million farmers and consumers who, so far, have few of the benefits of
advanced rice technology. The rainfed rice ecosystems share one major
characteristic: an uncertain moisture supply. This uncertainty is the

6 Rice Breeding and Genetics: Research Priorities and Challenges
to risk could be minimized by making cultivais available, more stable
yields and increase in productivity of resources. One way to improve
the well-being of rice-farming families in the rainfed lowlands is to
intensify the system by adding another crop such as a coarse grain or
foodgrain legume before and after rice, wherever this is possible.
Inland swamps have great potential in sub-Sahara Africa. About 138
million ha of these wetlands remain untapped in tropical Africa. About
1.7 million ha are cultivated with either rainfed or irrigated rice. It is
estimated that between 10 to 20 million ha of inland swamps are located
in West Africa. Emphasis on swampland development is a sound
approach for agricultural exploitation. Swamp rice farming is more
stable and productive than upland. In developed swamps, more than
one crop of rice or multiple crops can be grown depending on water
availability, Rice yield can be substantially increased with appropriate
technology. Recent experience has demonstrated that inland swamps
can be productive even with small-scale developments, since investment
in irrigation and drainage structures and management is minimal.
Centres of the Consultative Group on International Agricultural
Research (CGIAR) as well as the National Agricultural Research Stations
(NARS) have made contributions to boosting yields under relatively
favourable conditions. Some progress has been made in improved
farming practices for crop establishment, nutrient management, onfarm
water collection and weed control. However, more efforts-are
needed to address problems with drought, flooding and soil under less
favourable inland valley swamp conditions.
About 18 million ha of potential deepwater and tidal wetlands in
South and South-East Asia are not utilized. Africa and Latin America
also have large areas of unused deepwater and tidal wetlands. The
contribution of these areas could be enormous if constraints such as
excess water, salinity and hazards to human health could be overcome.
At present, deepwater rice is grown on 11% of the cultivated land,
which corresponds to about 16 million ha.
Large areas of floodplains and mangroves have been converted to
agricultural land. Cultivation of deepwater rice will remain an
important component in food production by rural people in the densely
populated floodplains and deltas of South-East Asia. Development of
wetland ecosystems such as mangroves and floodplains should not be
limited to the growing of rice, but integrated with other land-use
systems, including agriculture, agroforestry, wildlife management,
game ranching and ecotourism. The possibility of hsh culture integrated
with deepwater rice cultivars grown in water more than 50 cm deep for
at least one month during the growing season has to be explored. Every
---

F. Riveros 7
culture technologies and to encourage farmers to adopt such
technologies.
About 19 million ha of rice lie in the uplands of Asia, Africa, Latin
America and the Caribbean. The annual 20 million tons of production
support millions of people, most of whom live at subsistence level. The
uplands are diverse and usually have poor or degraded soils.
Topography ranges from sloping terraces to well-drained flatland. Rice
yields average 1 1 ha“^or less. Upland farmers are the poorest among the
world's rice farmers. 'Slash and Burn' agriculture leads to serious soil
erosion and degradation. This is further aggravated by shortening of the
fallow periods.
Serious efforts are needed in the upland ecosystem to arrest further
degradation of soil through diversification of rice-based cropping
systems. Mixed cropping of upland rice and pasture crops, for example,
has proven to be important for renovation of degraded pastures of
savanna in Latin America. The conservation of natural resources with
focus on the maintenance of soil and water conservation should be the
major objective in this agrbecology.
The United Nations Conference on Environment and Development
(UNCED) showed increasing concern about greenhouse emissions and
the conservation of biodiversity. Methane is an important greenhouse
gas: the concentration of methane in the atmosphere is currently
increasing at the rate of about 1% per year. Flooded rice fields are an
important source of methane on a global scale, contributing
approximately 25% of the total atmospheric emission, and mitigation
technologies are required to stabilize atmospheric methane
concentration in the long term. Possible mitigation technologies include
reducing inputs of easily degradable carbon; increasing soil and plantmediated
methane oxidation; reducing emission pathways through the
selection and breeding of rice cultivars; and preventing or reducing
methane formation through intermittent aeration, sources and mode of
fertilizer and the application of chemical inhibitors.
The development, adoption and impact of agricultural technologies
are not simply the business of scientists and farmers; they are indeed a
transect of the socioeconomic domain of a political system wherein
those who support research and development (R & D) as well as those
who benefit from it have a vested interest. An integrated approach to the
rice production system is mandatory.
In December 1993, the Uruguay Round of Multilateral Trade Negotiations
came to a successful conclusion. For agriculture, the basic aim of
the Uruguay Round was to provide substantial progressive reductions
in agricultural support and protection over an agreed period of time in
_________ 1. _____1 _____ ] ______

1
8 Rice Breeding and Genetics: Research Priorities and Challenges
agricultural markets. Overall the world rice market is likely to benefit
from the agreement. Rice quality will be given more serious
consideration in future.
The race between the demand for rice and production has so far
been evenly matched. We have not lost the race but the paradox still
exists. Millions of people still go hungry due to lack of access of food or
deficiencies in food distribution systems. The global population is
expanding each year at an alarming rate. More people must be fed with
food produced on less and less land and with less water and less labour.
The challenge for the future is tremendous: the demand for rice must
match with the population explosion and more rice must be produced
by increasing the productivity of land through environment-friendly
technology and optimum resource management. There is need to
diversify rice-based farming system to generate employment and moré
income for the poor farming community so as to arrest their migration
to urban areas. Apart from increasing the rice production, the major
thrust should also be to conserve the natural resources, soil and water as
well as biodiversity. Future generations should not pay a price for our
present exploitation of nature's endowments. The goal should be: 'Let
there be food for everyone. Let us strive to reduce hunger'.

Rice Breeding and Genetics:
Research Perspectives
J. S. Nanda*
THE CHALLENGE
The United Nation's recent population projections indicate that each
year almost 80 million people are likely to be added to the world's
population during the next quarter century. The world population
would increase by 35%, from 5,69 billion in 1995 to 7.67 billion by 2020.
The population increase will be more than 95% in developing countries,
whose share of global population is projected to increase from 79% in
1995 to 84% in 2020. Over this period, the absolute population increase
will be highest in Asia, but the relative increase will be greatest in sub-
Saharan Africa, where the population is expected to almost double
(Pinstrup Andersen and Pandya-Lorch, and Rosegrant 1997).
Demand for food depends on the population growth, its movement,
income levels, economic growth, human resource development, life
styles and preferences. Urbanization will contribute to changes in the
types of food demanded. The urban population in the developing world
is projected to double over the next quarter century to 3.6 billion . This
will profoundly affect the dietary and food demand pattern because of
the increasing opportunity cost of women's time, changes in food
preferences, changing life styles, and changes in relative prices
* Former Rice Breeding and Genetics Specialist, Guy/91/001, FAO of the United Nations,
Rome; Consultant, Palashoaili, Bhubaneswar. Oriasa. Tnriia.

The magnitude in growth and demand for rice will require new and
challenging research approaches to meet the food needs. Rice is grown
under four eco-systems, broadly defined on the basis of water regime:
irrigated, rainfed lowland, upland and flood prone. These four ecologies
account for 76,17, 4 and 3% of the current rice production respectively.
In order to raise the current level of production, Scobie et ah ( FAO, 1993)
suggested three areas which were to be sharply focused.
> Raising the yield frontier of rice which has not increased since IR8
was released.
> Sustaining the current yields, particularly of the intensive
irrigated systems which have shown accumulative stresses and
decline in yield.
> Closing the gap between potential yields and those achieved in
farming systems, particularly in the rainfed systems.
It is estimated that of the extra output of rice required by the year
2030,91.3% will be needed in Asia. Of that increase, 70% will have to be
10 Rice Breeding and Genetics; Research Priorities and Challenges
associated with rural-urban migration. The rural-urban migration leads
to diversified diets with a shift from basic staples such as sorghum,
millet, and maize to rice, wheat, livestock products, fruits, vegetables
and processed foods.
Prospects for economic growth during the next quarter century
appear encouraging, with global income growth projected to average
2.7% per year between 1993-2020. However, disparities in income levels
and growth rates both between and within countries are likely to persist
and poverty will remain entrenched in South Asia and Latin America
and to increase considerably in sub-Saharan Africa, unless these
countries implement perceptible and revolutionizing innovations.
Rice is one of the major cereals of the world and is the staple food for
about 2.7 billion people in Asia alone. Global demand for rice is
projected to grow at least equal to population growth, thus requiring a
70% increase in supply by the year 2025 (IRRI, 1993). The annual growth
rate in global rice production was only 1,8% during the 1985-92 period
compared to 2.8% during 1975-85 and 3.6% during 1965-75. Over the
years there has been a gradual decline in the armual growth rate of
global rice production. The population in rice consuming countries is
still growing at the rate of 1.8% per year. In order to meet the future
demand of rice, its production must be increased to at least match the
rate of increase in population growth to maintain the food-population
balance.
RESEARCH PRIORITIES

J.S, Nanda 11
produced in the irrigated systems, 21% in the rainfed lowlands, 6.3% in
the uplands and 3% in the flood-prone systems (FAO, 1993).
The target for the intensively cropped irrigated system is to raise the
experimental yield ceiling from 10 to 15 t ha"^ under tropical conditions.
In less favorable rainfed areas, the target is to increase and stabilize
yields at 2.5 to 3 t ha“^from 1 t ha“\
To make a quantum jump in the yield potential, it is imperative to
understand in depth the fundamental physiological processes that
determine growth and yield. The driving forces for crop yield are well
known; what is needed are both a source of nitrogen and carbon as well
as a sink for grain storage (Fischer, 1996). The opportunities to change
these processes are:
> Increasing the sink size and source through higher nutrient
assimilation, particularly N and higher carbon assimilation
>• Allocation of more storage reserves for grain
> Increasing net carbon assimilation during grain filling through
• enhancing photosynthetic carbon assimilation
• erdiancing light interception and canopy traits
• utilization of organic and inorganic carbon from soil
• lodging tolerance
• reduction of photorespiration and
• reduction in maintenance respiration.
> Enhancing the grain-filling duration.
Concurrent with transcending the present yield barriers, the yield
levels so achieved must be stabilized, providing durable resistance/
tolerance to biological and physical stresses through diversity in genes,
cultivars and species. The environmental component must be
appropriately addressed, ensuring the conservation and permanency of
the resource base of the intensive irrigated systems and how endogenous
biological processes can enhance the environment for the management
of pests and inputs (including nitrogen fixation) and minimize soil
losses in the uplands (Fischer, 1996) examined.
Irrigated Rice
The target yield potential in the irrgated system is 15 t ha"^ To meet this
target, the IRRI scientists proposed modifications of the present highyielding,
semidwarf plant types and developed a new plant ideotype for
direct seeded crop establishment.The traits targeted for this new plant
type are listed in Table 1.1 (Peng et al, 1994). Considerable progress has
been made in developing varieties with the new traits which may break
the yield barriers significantly.

Hybrid rice offers yet another opportunity to boost the yield
potential of rice. Hybrid rice has a yield advantage of 15-20% over the
conventional high yielding varieties (Virmani et al, 1993). There is
evidence that the level of heterosis may be further enhanced using
indica and tropical japónica hybrids based on the new plant type
germplasm. The hybrid technology is targeted for areas with a high
proportion of irrigated ecology and areas with a high-labor land ratio.
Heterosis in rice is exploited to a considerable extent in China. The high
productivity of hybrid rice enabled China to reduce its rice area from
about 34.4 Mha in 1978 to about 31.98 Mha in 1988 and at the same time
increase, its rice production from 136 Mt to 169.1 Mt during the same
period (Tran and Nguyen, 1998). This reduction in area of planted rice
not only promoted diversification in rice-based production systems for
more incomes and risk reduction, but also helped to minimize the
country's global emission of green-house gases such as methane and
nitrite oxide to the environment. Outside China, in hybrid rice was
planted about 11,000 ha in 1992,34,000 ha in 1993, and 102,000 ha in 1996
in Vietnam, with the average yield of 6 .5 1 ha“^or 15-30% higher than the
best commercial varieties. In India, farmers e:rew about 65,000 ha of
12 Rice Breeding and Genetics: Research Priorities and Challenges
Table 1.1
Traits of traditional and semidwarf rice varieties relative to the new
plant type under development at IRRI.
Plant Traditional Semidwarf, New plant type.
part/trait tall variety modern proposed
HYV 1960-1970 traits 1990
Height (cm) >120-150 90-110 90-110
Leaves Long, droopy Short, small, erect Thick, short, small,
erect
Tillers Low tillering Upright (compact). No unproductive
high tillering. tillers
Culm Tall and thin Short and stiff Short and stiff
Panicles 12-15/plant 15/plant 8/plant
Grains per panicle 90-100 80-100 200-250
Harvest index 0.30 0.50-0.55 0.55-0.60
Growth duration 160-200 110-140 100-130
(days)
Grain yield 3-4 t ha'^ (not 6-10 t ha'^ 10-13 t ha"^
potential N responsive) (N responsive) (N responsive)
Root system Vigorous Vigorous
Pests and diseases Variable resistance Multiple resistance Multiple resistance
Crop establishment Direct seeded or Direct seeded and Mainly direct seeded
transplanted transplanted
Varietal examples Peta Taichung
native 1, IRS
IR65598-112-2
Source: Fischer, 1996.

J.S. Nanda 13
rice was also reported from Bangladesh, Korea DPR and Myanmar.
However, its impact in other countries is yet to be felt. Considerable
research efforts are needed to enhance the level of heterosis, diversify
CMS source, boost seed yield, improve grain quality and resistance to
insect pests and diseases. Exploitation of apomixis will be a significant
step forward in resolving the bottle-neck in hybrid seed technology.
There are growing concerns world over about the extent, rate and
effects of degradation of natural resources such as soil, water and
biodiversity due to the demands of Increasing population to boost food
production. The irrigated rice culture has led to a build-up of salinity
and waterlogging, micronutrient deficiencies, formation of hard pan
and increased pest build-up, and plateauing in the yield of rice varieties.
Besides, the availability of water for irrigated rice is gradually being
reduced due to clogging of waterways, growing competition for water
betvyeen sectors, rising costs of developing new water resources. In
years to come, the irrigated ecology will continue to provide the greatest
potential to increase rice production. Sustainable rice production in the
irrigated ecology will be increasingly dependent on yield increase and
cropping intensity, as there is limited scope for the expansion of the net
area. More rice needs to be produced with less water and less
agrochemicals through efficient water management, integrated nutrients
and pest management,
Rainfed Rice
Rainfed rice systems (rainfed lowland, upland and flood prone) have
one major feature: uncertain moisture supply. Fields may have too
much water, leading to submergence, or exposing the crop to drought
stress. In the season the fields may be exposed to both drought and
submergence for variable periods of crop growth. Most of the world^s
resource-poor farmers grow rice in this risk-prone-ecology. It is
imperative to take measures to minimize the risk of farmers and enhance
their productivity.
Upland rice is grown on about 20.4 Mha, which represents about
14% of the world's rice area. Numerous subsistence farmers grow
upland rice mostly on poor, well-drained soil with an erratic rainfall and
under shifting or permanent cultivation, or as a pioneer crop. Rice is
grown either as a monocrop or in crop mixtures. The average yield in
this ecology varies from 1 to 1.5 t ha^. These areas, marginal for rice
production, are expected to grow rice in the medium term as long as the
production from lowlands fails to meet the demand. The major problems

14 Rice Breeding and Genetics: Research Priorities and Challenges
soil erosion, soil acidity, deficiency of micronutrients and weed
infestation. The principal measures for improving upland rice cropping
systems are water and soil conservation. Research should focus on
understanding the fundamental physiological features of upland rice.
New approaches are needed, including national policy reorientation
and political, will, in order to stabilize and reduce vulnerable upland
rice areas and make them more economic, productive, and sustainable
when exploited (Tran, 1986).
Rainfed lowland rice, including deepwater and tidal wetlands
constitutes about 31% of the world's harvested rice areas and 21% of the
world's rice production (IRRI, 1993). The yield potential of the shallow
rainfed lowland with assured rainfall is as good as that of irrigated
areas, but has remained unexploited. These areas provide great
opportunities for increased rice production. Woopereis (1993) suggested
that to enhance productivity in the rainfed environment, research should
focus on understanding processes and mechanisms to establish a basis
for developing site-specific applications, and appropriate crop models
are needed to stimulate growth and examine the effects of variable
weather on yields. Abiotic stresses such as submergence, elongation,
salt tolerance, P and, Zn deficiency tolerance, Al and, Fe toxicity
tolerance, drought, and cold tolerance are of significant importance in
rainfed ecology. Research should focus on developing reliable screening
techniques, indentifying suitable donors, and proper understanding of
genetic basis and inheritance.
BIOTECHNOLOGY
Application of biotechnology in crop improvement has tremendous
potential. It is estimated that by the year 2000, annual farm sales of
biotechnology-derived products are likely to total sothe $10 billion, with
70% based on seeds and 30% on veterinary products. Some $1 billion is
spent armually on global research and development in biotechnology.
Most of the bioengineering research carried out for cleveloping countries
to date has been to lay the groundwork for future crop transformation,
but improvements in crop yields could come rapidly. The Rockefeller
Foundation's support for rice biotechnology should begin to pay off in
two to five years in the form of new varieties available to some Asian
farmers. It is likely that efforts to improve the rice yield in Asia through
biotechnology will result in a production increase of lO to 25% over the
next 10 years (CGIAR,1997). The application of biotechnology in crop

J.S. Nanda 15
anthers or microspores (haploid breeding), exploitation of
somoclonal variation in plants regenerated from cultured cells,
protoclonal variation in protoplast-derived plants and their seed
progeny.
> Production of transgenic plant through DNA transfer to plant
protoplasts using either chemical treatment (polyethylene glycol,
PEG ) or electroporation for transformation.
> Insertion of genes coding for protein toxins from entomocidal
bacteria, protein inhibitors of insect digestive enzymes and
certain lectins for insect pest resistance, developing herbicideresistant
cutivars.
>■ Using molecular probes for disease diagnosis and monitoring,
which will allow more effective use of existing sources of
resistance.
V Developing rice genetic maps and markers to help identify the
most important genetic components, modifying the level of
expression of stress-induced rice genes and the transfer of alien
genes that enhance a desired response. The high density
molecular genetic map is of great value in map based cloning of
agriculturally important genes. Wang et ah (1995) used BAG
(bacterial artificial chromosome) libraries and identified clones
carrying gene for bacterial blight resistance. Song et ah (1995)
isolated by positioning cloning. The isolated gene was
introduced into several elite rice cultivars through transformation
(Khush et ah, 1998).
>* Protoplast fusion procedure, coupled with the regeneration of
plants from products of the fusion for the introgression of both
nuclear and cytoplasmic genes.
>• Introduction of symbiotic biological nitrogen fixation.
The greatest promise in rice improvement from the application of
biotechnology may stem from DNA-marker based gene identification
and gene insertion to create transgenic rice (Wu, 1994). Marker assisted
selection (MAS) was successfully employed for pyramiding four
different genes (Xa^ xa^, and Xa^^) for bacterial blight resistance
(Huang et ah, 1997). Transgenic rice plants can be produced harboririg
one or more insect resistant gene(s) (Table 1.2). Among the tested
insecticidal proteins, the four groups which are potentially useful
against rice insect pests are:
> Protease inhibitors such as serine proteases, cysteine proteases,
zinc proteases, and aspartyl proteases (Hilder et ah, 1993).
> “ Bacillus thuringiensis insecticidal proteins: A total of 20 insecticidal

16 Rice Breeding and Genetics; Research Priorities and Challenges
Table 1,2
Some examples of transgenic rice plants carrying agronomically
important genes
Transgene Gene transfer method Useful trait Reference
bar
Microprojectile Tolerance to herbicide Caocf fll., 1992
bombardment
bar PEG-mediated Tolerance to herbicide Datta et al., 1992
Coat protein Protoplast Tolerance to stripe virus Hayakawa
gene electroporation etal.,1992
Chitinase PEG-mediated Sheath blight resistance Lin et al., 1995
cnfIA(b) Protoplast Resistance to striped Fujimoto
electroporation stem borer et al, 1993
crylAib) Particle bombardment Resistance to yellow Wuhn et al, 1996
stem borer and striped
stem borer
crylAib) Particle bombardment Resistance to yellow
stem borer and striped
Ghareyazie
ei al, 1997
stembordr
crylAic) Particle bombardment Resistance to yellow
stem borer
Nayak
et al, 1997
CpTi PEG-mediated Resistance to striped Xu et al, 1996
stem borer and pink
stem borer
Com Protoplast Insecticidal activity Irie et al, 1996
cystatin (CC) electroporation for Sitophilus zeamais
Source: Khushetal (1998).
two major rice insect pests, striped stem borer and leaf folder,
than the untransformed control plants (Fujimoto et at, 1993).
> Lectins-Snowdrop lectin (GNA); GNA added to artificial diet has
been shown to inhibit the growth of brown planthopper (BPH).
The gene encoding GNA has been cloned (Boulter et al, 1993;
Gatehouse et fll.,1994). Once the GNA gene is introduced into
transgenic plants, it is expected that the plants will become
resistant to BPH and green leaf folder.
> Ribosome-inactivating proteins: This group of proteins has been
shown to inhibit the growth of certain species of insects and fungi.
The major fungal diseases which attack rice plants are Pyricularia
oryzae and Rhizocionia solani causing serious crop damage in the form of
blast and sheath blight diseases. The following proteins are potentially
useful against fungal pathogens of rice.
> Chitinase and p-l,3-glucanases: These enzymes are found in
plants arid microbes and are capable of degrading the cell walls of
fungi. In some instances, the antifungal activity of chitinase in
vitro is enhanced when applied in combination with p-1,3-
glucanase. Rice has been transformed using the chitinase sene

J.S. Nanda
and such transformed plants showed some resistance to R. solani
(Anuratha et al., 1994; Lamb et al.j 1994).
> Ribosome-inactivating proteins (RIP): Several genes coding for
RIP have been cloned. Rice has been transformed with a RIP gene,
and relatively high levels of RIP were found in transgenic plants.
The effectiveness of these transgenic plants against rice fungal
pathogens, including R. solani are under investigation.
> Thionins; Thionins are polypeptides with antifungal activities.
The usefulness of producing transgenic plants containing a
thionin gene needs to be explored.
> Antifungal peptides.
Wu (1994) has summarized (Table 1.3) the genes currently available
for producing insect resistant and fungal disease resistant transgenic rice
plants. He has projected that in about 12 to 18 years transgenic disease and
insect resistant rice varieties will be available to farmers. An estimated
annual benefit of using insect resistant and disease resistant transgenic
rice plants in the field is projected to be $ 13.4 billion (Table 1.4).
Table 1.3
Potential for producing insect and fungal resistant transgenic
rice plants (Wu, 1994)
Desired
new traits
Insect resistance
Fungal diseases
Target insect or fungus
Yellow stem borer; Striped
stem borer; Rice leaf folder.
Gall midge
Brown planthopper; Green
leaf folder
Sheath blight (Rsaloni)
Blast (P.oryzae)
Potentially useful genes
for transforming rice
Genes encoding protease
inhibitors CpTi, Pinll, SbTv, B.T.
genes: cryIA(b),crylA(c)f
Crylll, Pinll RIP z^nes
GÑA gene, RIP genes
Gene encoding chitinases,
fil, 3-glucanases, RIPs, thionins,
and antifungal peptides
Table 1.4 Estimates of effects or benefits of the rice biotechnology programs (Wu, 1994).
Trait or yield
enhancement
Multiple insect
resistance
Multiple disease
resistance
Time to production
(years)
Annual effect or benefit
after realization
Optinustic Conservative Area Yield Quantity Value in
(M.ha) (%) (Mt) billion
dollars
12 21
15 22
37 30 41 8.0
50 15 27 5.4

18 Rice Breeding and Genetics: Research Priorities and Challenges
> CP (coat protein) genes for the two viruses that cause tungro
disease have been cloned (Hay et ah, 1991) and efforts are underway
to express these genes in rice plants. A coat protein gene for
rice stripe virus was introduced into two japónica varieties by
electroporation of protoplasts (Hayakawa et ah, 1992). The
resultant transgenic plants expressed high levels of CP and
exhibited a significant level of resistance to virus infection; this
resistance was inherited by the progeny.
> Starch levels and dry matter accumulation were enhanced in
potato tubers of plants transformed with glgc^^ gene from E. coli
encoding ADPGPP (ADP-glucose pyrophosphorylase), the critical
enzyme for inregulating starch biosynthesis in plant tissues (Stark
et ah, 1992). The glgc^^. gene has been introduced into rice at IRRI
and its expression is being investigated (Khush et ah, 1998).
The advent of molecular markers is expected to modify rice breeding
drastically^ since they enable the number of basic progenitors to be
narrowed down and allow the creation of rice genetic maps.
Biotechnological approaches need a high investment. It is imperative to
think a balance between the conventional and biotechnological
approaches. They should be complementary rather than competitive for
scarce resources.
THE VISION
Rice production should not be viewed in isolation but as a part of a
holistic farming system in which the farmer's income and welfare as
well as the diversity of the social biological and physical environments
should be integrated into the design of appropriate technologies.
Technological advancements are useful when accompanied by
appropriate national policies which are supported by consistent and
concrete programs. Therefore, long-term sustainable rice production
requires the formulation and implementation of relevant program for
rice research, development and production (Shastry ei fll.,1996).
The 2020 vision is of a world 'where every person has access to
sufficient food to sustain a healthy, and productive life, where
malnutrition is absent, and where food originates from efficient,
effective, and low-cost food systems that are compatible with
sustainable use of natural resources' (IFPRI, 1995). Existing technology
and knowledge will not permit production of all the food needed in 2020
and beyond. Most of the increase in food production will have to come

J.S. Nanda 19
economically or environmentally sound option in most parts of the
world. Some of the yield increase will occur as more inputs are used and
as production methods are improved. However^ accelerated investment
in agricultural research is essential to achieve the required productivity
increases. Low-income developing countries are grossly under investing
in agricultural research compared with industrialized countries, even
though agriculture accounts for a much larger share of their employment
and incomes. Their public sector expenditures on agricultural research
are typically less than 0.5% of the agricultural gross domestic product,
compared with about 1% in higher income developing countries and
2-5% in industrialized countries (Pardey et at, 1991). There is also a need
to expand and realign international development assistance. Developed
coimtries had agreed to allocate at least 0.7% of the gross national
product (GNP) to foreign assistance. Most countries have not reached
this target and reduced their average contribution to 0.3% of GNP. The
current downward trend in international development assistance must
be reversed. To improve the effectiveness of the aid, each recipient
country should develop a coherent strategy for achieving its goals
related to food security, poverty, and natural resources and should
identify the most appropriate uses of international assistance (Pinstrup-
Andersen and Pandya-Lorch, 1996),
References
Anuratha,C.S., Lin,W., Muthukrishnan, S., Datta, S., Potrykus, L, Vidyasekarati, P. and Mew,
T. 1994, Genetic engineering of rice for resistance to sheath blight and insects , 7th
Meeting Inti. Program on Rice Biotechnology, Abstract, p.l59.
Boulter, D., Gateouse, A.M.R. and Wang, M. B. 1993, Genetically engineered insect resistance
to the brown plant hopper (BPH) and other sucking insects, 6th Annual Meeting Inti,
Program on Rice Biotechnology, Abstract, p. 9.
Cao, J., Duan, X., McElroy, D.M. and Wu, R. 1992, Generation of herbicide resistant
transgenic rice plants following microprojectile-mediated transformation of suspension
culture cells. Plant Cell Rep., 11:586-591.
CGIAR( Consultative Group on International Agricultural Research). 1997. Press Release, 9
Oct., 1997, Washit^toni.DC.
Cocking, E.C. 1996. Progress in rice biotechnology. In: Proc. 18th Session Inti, Rice Comm.,
Rome, Italy 5-9 September, 1994. FAO, Rome, pp. 39-43.
Datta, S.K., Datta, K,, Soltanifar, N., Donn, G. and Potrykus. 1992. Herbicide resistant indica
rice plants from IRRI breeding line IR72 after PEG mediated transformation of
protoplasts. Plant Mol, Biel., 20:619-629.
FAO, 1993. Investment in rice research in the CGMR : a global perspective. Report of the
Inter-Cehtre Review of Rice., AGR/TAC:IAR/93/4, Technical Advisory Secretariat,
Rome, 84 pp.

20 Rice Breeding and Genetics: Research Priorities and Challenges
Fujimoto, H., Itoh,K., Yamamoto, M., Kyozuka, J. and Shimamoto, K. 1993. Insect resistance
rice generated by introduction of a modified gamma-endotoxin gene of Bacillus
thuringiensis. Biotechnology, 11:1151-1155.
Gatehouse, J. A ., Powell, K., Wang, M., Hilder, V. and Boulter, D. 1994, Progress towards
transgenic rice with resistance to rice brown planthopper. 7th Meeting Inti. Program on
Rice Biotechnology, Abstract, p. 59.
Ghareyazie, B., Alinia, F., Menguito, C.A., Rubia, L., de Palma, J.M., Liwanag, E.A., Cohen,
M. B., Khush, G.S. and Bennett, J. 1997. Enhanced resistance to two stem borers in an
aromatic rice containing a synthetic crylAtb) gene. Mol. Breed. 3:401-414.
Hay, J.M., Jones, M.C., Blakebrough, M., Dasgupta, I., Davies, J. and Hull, R. 1991. An
analysis of the sequence of an infectious clone of rice tungro bacilliform virus, a plant
pararetovirus. Nucleic Acids Res. 19:2615-2621.
Hayakawa, T., Zhu, Y., Itoh, K. and Kimura, Y. 1992. Genetically engineered rice resistance
to rice stripe virus, an insect-transmitted virus. Proc. Natl. Acad. Set. USA 89:9865-9869.
Hilder, V. A., Gatehouse, A.M.R. and Boulter, D, 1993. Transgenic plants conferring insect
resistance: protease inhibitor approach. In; Transgenic Plants S.D. Kung, and R. Wu, (eds.)
Acad. Press, NY, voi. l, pp. 317-338.
Huang, N,, Angekes, E.R., Domingo, J., Magpantay, G., Singh, S., Zhang, G., Kumaravadivel,
N. , Bennett, J. and Khush, G.S. 1997. Pyramiding of bacterial blight resistance genes in
rice: marker assisted selection using RFLP and PCR. Theor. Appl. Genet. 95:313-320.
IFPRI (International Food Policy Research Institute). 1995, A 2020 vision for Pood
Agriculture and the Environment ; The vision, challenge and recommended action.
IFPRI, Washington, DC, USA, 50 pp.
Irie, K., Hosoyama, H., Takeuchi, T., Iwabuchi, K,, Watnabe, H., Abe, M. and Arei, S. 1996.
Transgenic rice established to express corn cystatin exhibits strong inhibitory activity
against gut proteinases. Plant Mol, Biol. 30:149-157.
IRRI (International Rice Research Institute). 1993.1993-1995: IRRI Rice Almanac, Los Banos,
Philippines.
Khush, G.S,, Bennett, J., Datta, S.K., Brar, D;S. and LI, Z, 1998. Advances in rice genetics and
biotechnology. Paper presented at 19th session Inti. Rice Comm., PAO, Cairo, Egypt, 7-9
September, 1998,
Lamb, C.J., Zhu, Q., Xu, Y., Dabi, T. and Nelson, A. 1994. Engineeringerüiahced.resistance to
microbial disease in rice, 7th Meeting Inti. Program on Rice Biotechnölögy^ Abstract,
p.l58.
'
Lin, W,, Anuratha, C.S^. Datta, K., PotrykUs, I., Muthukrishnan, S. and Datt^/ S.BÎ. 1995.
Genetic engineering of rice for resistance to sheath blight. Biotechnology 13:686-691.
Nayak, P., Basu, D., Das, S., Basu, A.^ Ghosh, D., Ramakrishnan, N.A., Ghpsh> M. and Sen,
S.K. 1997. Transgenic elite indica rice plants expressing CryIA(c) Ä-endotoxin of Bacillus
thuringiensis are resistant against yellow stem borer (Sciropophaga incertuldsy: Proc. Natl.
Acad. Sei. USA 94: 2111-2116. ,.}
Pardey, P.G., Roseboom, J. and Andersen, J.R. (eds.) 1991. Agricutural Research Policy:
International Quantitative Perspectives. Cambridge, U.K, Cambridge University Press:
462 pp. ■
Peng, S., Khush,G.S. and Cassman, K.G. 1994. Evolution of the new plant ideotype for
increased yield potential. In; Breaking the Yield Barriers: K.G.Cassman (ed) Proc. Workshop
on rice yield potential in favourable environments. Chap. 2. Los Banos, Philippines, IRRI.
Pinstrup-Andersen, P., Pandya-Lorch, R. and Rosegrant, W.1996. The world fdod situation:
recent developments, emerging issues and long term prospects, 2020 Visión Pôod Policy
Report, International Pood Policy Research Institute, Washington, D.C.. d. 36.

J.S. Nanda 21
Shastry, S.V., Trariy D.V., Nguyen, V.N- and Nanda, J.S. 1996. Sustainable integrated ríce
production. In; Proc. of the 18th Session Inti. Rice Comm. FAO, Rome, pp. 45-58.
Song, W. Y., Wang, G.L., Chen, L.L., Kim, H.S., Holsten, T., Wang, B., Zhai, W.X., Zhu, L.H.,
Fauquet, C. and Ronald, P. 1995. The rice disease resistance gene, Xa21, encodes a
receptor kinase-like protein. Science: 270:1804-1806.
Stark, D.M., Timmerman, K.P., Barry, G.F., Preiss, J. and Kishore, G.M. 1992. Regulation of
the amount of starch in plant tissues by ADP glucose pyrophosphorylase. Science: 285:
287-292.
Tran, D.V. 1986. An overview of the upland rice in the world. In; 'Progress in Upland Rice
Research', pp. 51-66, Manila, IRRI.
Tran, D.V. and Nguyen, V.N. 1998. Global hybrid rice: Progress, issues and challenges. Paper
presented 19th Session Inti. Rice Comm,, FAO, Cairo, Egypt, 7-9th September 1998.
Virmani, S.S., Prasad, M.N. and Kumar, J. 1993. Breaking the yield barrier of rice throughout
exploitation of heterosis. In: K. Murailidharanand E.A. Siddiq (eds.) New Frontiers in Rice
Research, Directorate of Rice Research, Hyderabad, India, pp. 76-85.
Wang,G., Holsten, T.E., Song, W.Y., Wang, H.P. and Ronald, P.C. 1995. Construction of a rice
bacterial artificial chromosome library and identification of clones linked to Xa^^ disease
resistance locus. Plant J. 7:525-533.
Woopereis, M.C.S. 1993. Quantifying the impact of soil and climate variability on rainfed
rice production. (Ph.D Thesis). Wageningen Agrie. Univ. The Netherlands. 188 pp.
Wu, R. 1994. Report of the committee on Genetic Engineering {Molecular Analysis of Rice
Genes). Rice Genetics News Letter.ll: 59-63.
Wuhn, J., Kloti, A., Burkhardt, P.K., GhoshBiswas, G.C., Launis, K„ Iglesias, V.A. and
Potrykus, 1.1996. Transgenic indica rice breeding line IR58 expressing a synthetic
gene from Bflci7/«s thuringiensis provides effective insect pest control. Bioiechonology
14: 171-176
Xu, D., Xue, Q., McElroy, D., Mawal, Y., Hilder, V.A. and Wu, R. 1996. Constitutive
expression of a cowpea trypsin inhibitor gene CpTi, in transgenic rice plants confers
resistance to two major insect pests. Mo/. Breed. 2:167-173.

Hybrid Rice
J. S. Nanda* and S. S. Virmani*^
INTRODUCTION
Exploitation of heterosis (hybrid vigor) in corn has been a landmark in
crop breeding. Heterosis has been exploited in a nuinber of crops such as
bajra^ sorghum^ cotton, sunflower etc. However, its use has been limited
in self-pollinated crops. In 1926, Jones first reported the occurrence of
heterosis in rice. Its commercial exploitation was demonstrated in China
in 1976 after development of the first set of stable and high yielding threeline
hybrid rice varieties by Professor Yuan Long Ping and his team of
scientists. The area planted to hybrid rice in China rapidly increased to
about 17 Mha in 1995. Hybrid rice varieties outyielded the commercial
check varieties by about 15-20%. The high productivity of hybrid rice
enabled China to reduce its rice area from about 34.4 Mha in 1978 to
about 31.98 Mha in 1988 and at the same time increased its rice
production from 136.9 Mt to 169.1 Mt during the same period (Tran and
Nguyen, 1998). Presently, China is the leading producer of hybrid rice in
the world (Table 2.1) and more than 50% of the 32 Mha of rice area is
under hybrid rice, accounting for more than 70% of the total production.
Outside China, hybrid rice was planted to about 11,000 ha in 1992,34,000
ha in 1993, and 102,000 ha in 1996 in Vietnam, with an average yield of 6.5
t ha'^ or 15-30% higher than the best commercial varieties (Tran and
Nguyen, 1998). In India, farmers grew about 65,000 ha of hybrid rice in
1996, Limited commercial cultivation of hybrid rice was reported in
Bangladesh, Korea DPR and Myanmar.
* Former Rice Breeding and Genetics Specialist GUY/91/001, FAO of the United Nations,

24. Rice Breeding and Genetics: Research Priorities and Challenges
Table 2,1
Area, production and yield of hybrid rice in China
Kind Area (Mha'^) Production (Mt) Yield (kgha'^)
Total rice 32,61 185.41 5685.70
Conventional rice 14.97 67.93 4531.50
Hybrid rice 17.64 117.48 6660.10
Hybrid rice (as of total rice) 54.10 63.36 117.10
Source: Xizhi and Mao (1994).
Heterosis is defined as the superior performance in growth, vigor,
vitality, reproductive capacity, stress resistance, adaptability, grain yield,
grain quality and other physiological traits of the Fj population of two
genetically diverse parents (P) compared to either the mid-parent (MP) or
better parent (BP) of the cross or to the check (CK). Heterosis is expressed
as:
I. Mid-parent heterosis (MP)
H ^ X 100%
MP
II, Heterobeltiosis or heterc^is over the better parent (BP) value
H = Fi"BP
BP
X 100%
III. Standard heterosis or heterosis over the check variety (CK)
p „ p v
H ^
CK
X 100%
In general, the expression of increased vigor of Fi hybrid over its
parents is called positive heterosis and that of decreased vigor is
designated as negative heterosis.
The success story of hybrid rice in China aroused spurts of interests at
the International Rice Research Institute (IRRI), Philippines and in many
national rice research programs to intensify research on hybrid rice. The
trials conducted at IRRI (Table 2.2 and 2.4), Philippines and in several
national programs, viz. Philippines, India, Vietnam, and Malaysia (Table
2.3), the rice hybrids outyielded the best check variety .(Virmani, 1996).
Many heterotic rice hybrids are released for commercial cultivation in
countries other than China (Table 2.5),
The positive standard heterosis in grain yield reported in rice
hybrids is attributed to the increased dry matter production due to
increased leaf area index, higher crop growth rate and harvest index,
due to high spikelet number and increased grain weight (Ponnuthurai et
'l 0R4V
in nnm VtPr rtf n a n irlta c « » r n la n f anilrtiliafc! niii-

J.S. Nanda and S.S. Virmani 25
Tabic 2.2 Comparison of the highest yielding rice hybrids and check
variety in yield trail at IRRI during 1986-95
Season Trial Hybrid
Yield
(t ha'b
Difference Percentage
of check
Growth
duration
(1) (2)
(3) (4) (5) (6) (7)
1986 DS I. IR54754A/IRR6R 7.4 1.2 119 126
n IR54754A/ARC 11353R 7.9 2.3 142 133
1986 WS I IR54752A/IR64 3.9 1.0* 134 126
n IR19728A/IR25167-9-2 3.6 0.6* 120 122
III IR46830A/IR5OR 4.1 0.7* 120 110
1987 DS I IR46830A/IR29723-143-3-2-IR 6.4 1.8* 139 112
II IR54755A/IR2797-125-3-3-2R 7.8 1.8* 130 130
III IR54752A/ARC 11353R 6.8 1.0* 117 128
1988 DS IV IR54752A/IR64R 5.3 0.9* 120 120
VI IR54752A/IR13146-45-2-3 6.8 1.1* 119 116
1988 WS IR46830A/IR9761-19-IR 3.2 1.0* 145 105
1989 DS I IR54752A/IR9761-19-IR 6,3 1.5* 131 111
II IR54752/IR28228-119-2-3-1-IR 6.5 1.0* 119 124
1989WS II IR54752A/54742-22-19-3R 3.5 0.8* 131 136
1990DS IV IR58025A/Ir29723-143-3-2-IR 5.6 1,0* 121 128
1991DS I IR62829A/IR35366-62-1-2-2-3 4.7 0.7* 118 112
II IR58025Á/IR54745-2-45-3-2-4R 5.4 1.2* 128 122
1991WS I IR58025A/IR19058-107-IR 6.4 1.2* 123 113
n 1R62829A/IR47310-94-4-3-IR 5.1 1.1* 128 120
1992DS III IR58025A/IR46R 6.3 0.8* 114 117
1992WS IV IR58025A/IR54056-64-2-2-2 4,4 0.7* 119 126
1993DS II IR58025A/IR34686-179-1-2-IR 7.4 0.8* 112 128
1993WS I IR58025A/BG915 4.0 0.8* 125 105
1994WS I IR58025A/IR59606-119-3 5.8 1.0* 121 105
II IR68275A/IR46R 6,6 1.6* 132 122
1995DS I IR58025A/IR58103-62-3 9,5 1.6* 120 114
II IR58025A/RP633-76-IR 9.6 1.6* 120 121
DS, Dry season; WS, Wet season.
* Significant at least at 5% level using LSD test.
Source: Virmani (1996),
compensation of yield components and cultivation practices (Kim, 1985;
Akita, 1988).
Rice hybrids showing positive heterosis for adverse temperature,
soil conditions and water regime will be of immense importance in
developing rice hybrids for stress environments. Tian et al. (1980) have
reported positive heterosis for drought tolerance, Akbar and Yabuno
(1975) and Senadhira and Virmani (1987) for salt tolerance; Chauhan
et al. (1983) for ratooning ability and Singh (1983) for deep water
tolerance in rice hybrids.

26 Rice Breeding and Genetics; Research Priorities and Challenges
Table 2.3
Yield advantage of some elite IRRI-bred rice hybrids in
national trials during 1990-1993
Country Year and
season
(1) (2)
India
Location
(3)
Hybrid
(4)
Yield Percentage
(tha'^) of best check
(5) (6)
1990 WS Mandya IR58025A/IR976M9-IR 9.3 112
1991 DS Mandya IR58025A/IR29723-143-3-2-IR 7.0 125
Hyderabad IR62829A/IR10198-66-2R 6.8 128
1991 WS Hyderabad IR62829A/IR10198-66-2R 7.0 117
Maruteru IR62829A/IR35366-40-3R 6.3 112
New Delhi IR62829A/IR35366-40-3R 11.6 123
Hyderabad IR62829A/IR28238-109-2R 5.2 121
Mandya 1R58025A/IR40750-82-3R 6.4 121
IR58025A/IR54742-22-19-3R 6.2 117
Faizabad IR58025A/IR40750’82-3R 4.6 126
1992 WS Hyderabad IR58025A/IR34686-179-1-2-1R 7.3 109
Mandya IR58025A/IR32419-28-3-1-3 7.5 114
Coimbatore IR58025A/IR39323-182-2-3-3R 6.3 131
Faizabad IR58025A/IR46R 6.6 135
Cuttack IR62829A/IR46R 5.3 143
Vietnam
1990 WS Omon IR62829A/IR29723-143-3-2-1R 6.7 129
Omon IR58025A/IR29723-143-3-2-1R 7.6 146
1991 DS Omon IR62629A/IR29723-143^3-2-lR 6.1 122
Omon IR58025A/IR29723-143-3-2-1R 6.0 120
1991 WS Omon IR62928A/IR29723-143-3-2-1R 4.9 132
Omon IR58025A/IR29723-143-3-2-1R 4.6 124
1992 WS Omon IR62829A/IR29723-143-3-2-1R 5.5 112
Omon IR58025A/IR21567-18-3R 6.2 112
Omon IR58025A/IR52287-15-2-3-2R 6.7 144
Hanoi IR58025A/IR29723-143R 5.1 150
IR58025A/IR66R 5.1 150
IR62829A/IR10198-66 5.7 167
1993 DS Omon IR62829A/IR47310-94-4-3-1R 5.8 110
Tanhiep IR62829A/IR47310-94-4-3-1R 7.3 112
Binhau IR58025A/IR54742-22-19-3R 7.5 101
Philippines
1992 DS Maligaya IR58025A/IR32419-28-3-13 7.9 113
San Mateo IR62829A/IR20933-68-21-1-2R 8.0 139
1993 DS Maligaya
Group I IR64608A/IR29723-143-3-2-1A 7.8 122.
Group II IR58025A/IR54742-22-79-3R 7.7 131
San Mateo
Group I IR64608A/IR29723-143-3-2-1R 6.3 102
Group II IR5025A/IR34686-179-1-2-1R 8.1 112
Malaysia
1990 DS Bumbong IR62829A/IR29723-143-3-2-1R 5.8 141
Lima IR58025A/IR29723-143-3-2-1R 5.1 124
1991WS Bumbong IR58025A/IR29723-143-3-2-1R 5.6 140
Lima
/nanrirzrio ida q o 11> K£. 117

J.S. Nanda and S.S. Virmani 27
Table 2.4
Yield advantage of some elite IRRI-bred rice hybrids evaluated in
some collaborating countries during 1993-94
Country Location Season
Hybrid
Yield
(tha-^)
Difference
from check
i
1
(1) (2)
(3)
(4)
(5)
(t ha'^)
(6)
1 • India
'i Hyderabad 1993WS IR58025/IR54742 6.4 1.1*
; IR58025A/IR32809 5.8 1.5*
IR58025A/IR13419 5.6 1.3*
i
IR58025A/IR72R 5.6 1.3*
{
IR62829A/IR40750 5.2 0,8*
■i 1994 DS IR62829A/IR40750 7.0 1.5*
1 IR58025A/IR21567 7.5 1.2*
n 1994 WS IR58025A/RP633-76-1 7.7 1.2*
1 IR62829A/IR29723-143R 7.7 1.2*
■i Karnal 1993 WS IR58025A/IR20933 9.6 0.9*
i IR58025A/IR10198 9.8 0.8*
1 Delhi 1994 WS IRa8025A/IR34686-179 5.2 1.0*
j Kapurthala 1993 WS IR62829A/IR29723 9.6 2.4*
1994 WS PMS10A/BR827-35-R 6.6 2.6*
Chinsura 1993 WS IR58025A/IR10198 5.6 1.0*
i 1993 WS IR62829A/IR54883 6.2 1.6*
I Coimbatore 1993 WS .IR58025A/IR32809 5.6 1.3*
j Wyra 1994 WS IR58025A/IR54791-19R 7.9 1.3*
i
i Mandya 1993 WS IR58025A/IR13419 6.6 1.3*
■] IR58025A/IR54742 7.1 2.4*
1 1994 DS IR58025A/RP1057 10.0 2.5*
IR62829A/IR40750 8.7 1.5*
PMS8A/IR46R 8.7 1.5*
1994 WS IR58025A/IR54969-41 9.2 2.8*
IR58025A/RP633-76-1 8.0 1.6*
■) Faizabad 1993 WS IR58025A/IR13419 7.6 1.5*
■1 1994 WS IR58025A/IR25912-81 7.3 1.7*
Karjat 1994 WS IR58025A/BR827-35R 3.2 2.5*
1 Pakistan
1 Kalashah Kaku 1994 WS IR58025A/IR21567 6.9 2.4*
1 Sialkot 1994 WS IR58025A/IR21567 5.1 1.2*
j Farookabad 1994 WS IR58025A/IR29723 6.4 0.6*
Philippines
Maligaya 1994 WS IR68275A/IR46R 6.5 1.1*
i San Mateo 1994 DS PMS8A/IR29723 8.7 1.1*
IRRI 1994 WS IR58025A/IR59606-119 5.8 1.0*
IR68275A/IR46R 6.6 1.3*
1\
n IR58025A/IR52774-B-B 6.6 1.3*
Vietnam
ij Omon 1994 DS IR58025A/IR34686 7.8 1.9*
'■i
-jt IR58025A/IR25912 7.3 1.3*

28 Rice Breeding and Genetics: Research Priorities and Challenges
Table 2.5
Heterotic rice hybrids released for commercial cultivation
during 1993-94 in countries other than China
Country Hybrid Released as Year Recommended for
Vietnam IR58025A/IR29723 UTL-1 1993 Mekong Delta
IR62829A/IR29723 UTL-2 1993 Mekong Delta
India IR58025A/Vajram APHR4 1994 Telangana and
Rayalseema
IR62829A/MTU9992 APHR-2 1994 Telangana and
Rayalseema
IR62829A/IR10198 MGR-1 1994 Tamil Nadu
IR58025A/IR9761-194R KRH-1 1994 Karnataka
IR58025A/KMR 3R KRH-2 1994 Karnataka
IR62829A/Ajaya R CHNRH-3 1996 Boro Season
in West Bengal
IR58025A/IR40750 DRRH-1 1996 Telangana and
Rayalaseema
Philippines ' IR62829A/IR29723-143-3-2-1R Rc26H
(Magat)
Source: Virmani (1996); Siddiq et ah (1996).
GENETIC BASIS OF HETEROSIS
1994 Cagayan
Valley
The genetic basis of heterosis is sought in finding explanations for
inbreeding depression in cross-pollinated crops and vigor expressed in
hybrids of inbreds. There are two hypotheses to explain heterosis:
(i) Dominance hypothesis.
(ii) Overdominance hypothesis.
Dominance Hypothesis
Davenport suggested the overdominance hypothesis in 1908. According
to this hypothesis^ heterosis results from action^ interaction and
complementation olfavorable dominant genes brought together in an
hybrid of two inbreds. The dominant genes are presumed to be favorable
and recessive genes deleterious for vigor and growth. If such an
explanation holds true to account for heterotic vigor in Fj hybrids^ it
should be possible to accumulate favorable genes in inbreds, which
should perform as well as the hybrid. However, the large number of
genes involved in quantitative characters such as in grain yield and
linkage of deleterious recessive genes with favorable dominant genes
preclude the possibility of recovering homozygous lines as vigorous as
the hybrid.

J.S. Nanda and S.S. Virmani 29
î î î ,
i-< m es
rH rH
T—1î
i-l rH 00
* t J
V ) (N
rH X~H
î î ,
CS O î
x-H x-H os
1
X
T i
.S
Ü
O *VO *'©
Ê *vo *a^ ’*00
*iN *. LO -S'
î *co
0 0 i-i o
*t-i *cs î
Ï-H X—Í vo
Í î
VO v:t< Y
* o î Í
x-l t>. CO
§
0
1 OJ c;
0
Oh
1
£
*o*
î î î
i-i 00 -rP
CO cr> fO
î î ■î
OV tW 1—l
tH (N es
á *co 'co *(S ' f Îh
Ê
Í t Í
o\ 00 00 ^tr> e s CO
î î î
(Jv r-4 ,-t
(S CO CO
Í Í î
Ov O' Os
î , î
T—1«
t—( Œv t-H
î î î
-Vi< 00 V s
tM tH cH
l“ C CO
1 1 1
î i î
m tn '(t*
£
ê
•Í3
en
P
g
o
T-H es T—Í
^ ^ ^
(N es ujl es es cc)
'•S S
I“ l T-1 CO (T) OJl rH rH cp rH o Y
TJ
•i3 VI
ê
-ê
? :ôQLi
(U
ê
-t
g
k > * , - .
Uj \£) (N
î î î
in o\
'X) LO (N
î ä *
i-< o Í
o o r i
t-H T—1 ÏS'
: î
o es lO
CS eo
Í î î
tH 00 CO
Tjl Tt'
‘l>^ *LO o
CS CO
Vo *Lo Vs
\ 0 ts. os
î î Í ,
S i ■n' CS
T-l (S CS
*'!l^ o *s#i
LO lO CO
t î î
x-( lO CO
in
i : î
os O X—1
(OS Os 00
t î î
CS vO O
rH
*00 ‘lo *co
CO CO CO
'xH CO Ov
(S CS es
î î Í
oo SP oo
VD v5 I>.
o "Ü* vo
£
in iN ^ eO rM LO ro c^ ^ x~i ^0
! ç-
-Ö
fin
S '
Ê
£
w 00 ep CO CS \o
1 r-<
CO o es
!
tW Tifi t s
1 1—'
tH X—1 ^ (H o r#
rH XH 1
vo (3s
T!l( UJl CJV
■ÏÏ K w
O nj
iJ

30 Rice Breeding and Genetics: Research Priorities and Challenges
homozygote for the same gene. If A1 and A2 are the two contrasting
alleles for a single locus, the hétérozygote combination A1A2 is more
favorable to the plant than either of the homozygote combinations A1A1
or A2A2, This phenomenon is called overdominance. Overdominance
has been demonstrated in several traits that are controlled by a single
gene (Berger, 1976). Heterosis can result from partial to complete
dominance, overdominance, epistasis and a combination of these factors
(Comstock and Robinson, 1952). If partial to complete dominance
predominates, it is theoretically possible to develop homozygotes with a
performance equal or superior to hybrids. However, if overdominance
type of epistasis predominates, then the highest-yielding lines must be
heterozygotes (Sprague and Ebehart, 1977). Jinks (1983) observed
inadequate evidence of genuine overdominance for quantitative
characters and opined that apparent dominance due to non-allelic
interaction and linkage disequilibrium was a common contributor to
heterosis. It is therefore argued that in such situations recombinant
inbred lines as high yielding as or superior to Fj hybrids can be
developed provided appropriate breeding schemes are followed. Jinks,
however, did not rule out the possibility of masking effects of a wide
range of incomplete or complete dominance at the individual loci
including overdominance at some. He concluded that such a situation
would merely result in loss of accuracy in the prediction of the range of
pure breeding families extractable from heterotic cross. The best of these
families would produce second cycle heterotic FjS, if they were crossed
to another family which carried alternative alleles at the loci displaying
overdominance. The genetic basis of the exploitation of heterotic rice
hybrids in China and elsewhere can be attributed to the aforesaid
phenomenon as explained by Jinks.
Although biometrical analysis gives some insight into the relative
importance of dominance and overdominance, such analyses are not
powerful enough to elucidate the role of overdominance if it exists only
for a reduced part of the set of loci in the quantitative traits of interest
(Gallis, 1988). It is difficult to establish conclusive proof for the genetic
basis of heterosis for either of the hypotheses proposed because of the
complexity of inheritance of quantitative characters such as yield. All
types of gene interactions both inter-and intra-allelic are probably
involved. Gallis (1988) stated that heterozygosity for regulatory systems
might be more significant for the explanation of heterosis. Heterozygosity
for regulatory genes may lead to greater homeostasis in a variable
environment and heterosis is the result of genotype x environment
interaction and such a mechanism will be a fundamental property at the
dioloid level. Gallis fl988) observed that in both auto- and allogamous

J.S. Nanda and S.S. Virmani 31
their studies observed a low correlation between marker heterozygosity
and trait expression, indicating that the overall heterozygosity made
little contribution to the heterosis. Their results provided strong
evidence that epistasis played a major role for the expression of
heterosis.
Heterosis is a phenomenon of superior growth, development, differentiation
and maturation caused by the interaction of genes, metabolism
and environment. Therefore, a logical approach to explain heterosis
would be to take into account the nuclear genome heterozygosity as well
as the effect of cytoplasm. Several studies in rice have demonstrated the
positive and negative effects of cytoplasm on agronomic as well as
physiological characters (Maruyama et at, 1985; Sasahara et ah, 1986;
Chen et ah, 1987; Young and Virmani, 1990). Wang and Wen (1995)
studied the effect of sterility inducing cytoplasm in seven isogeneic
alloplasmic male sterile lines and their maintainer, Lu-Hoongzao IB.
They observed significant positive effects on spikelets/panicles (CMS-K
and CMS-WA cytoplasm); significant effects on grain yield/panicle
(CMS-L, CMS-J, and CMS-WA cytoplasm); seed set percentage (CMS-L,
CMS-J, CMS-Y, CMS-S, CMS-D, CMS-K and CMS-WA cytoplasm); 1000
grain weight (CMS-I, CMS-J, CMS-Y, CMS-S and CMS-WA cytoplasm)
and spikelets/panicle (CMS-J cytoplasm).
PREDICTION OF HETEROSIS
Breeders are always interested in choosing parental lines which will
result in a heterotic combination without making all the crosses possible
among the potential parents. Several methods such as per se performance
of the parents, genetic diversity determined through geographic origin,
multivariate analysis using morphological and agronomic traits, isozyme
and restriction fragment length polymorphism (RFLP) and combining
ability armlysis have been used to predict the best possible heterotic
parental combination.
In the hybrid rice breeding program's several internationally known
commercial rice varieties and elite breeding lines have been used such as
IR24, IR26, Milyang46, IR661, IR9761-19-1R. Therefore, per se performance
is a basis for selecting parental lines for developing heterotic
hybrids. However, it is also important that the parents selected to
develop heterotic hybrids are adapted to the prevailing conditions for
which hybrids are developed.
The genetic basis of heterosis lies primarily in the inter allelic and/
or intra-allelic genetic differences among the parents. Therefore,

32 Rice Breeding and Genetics: Research Priorities and Challenges
be related to the geographic origin of the parental lines. However,
internationalization of plant breeding efforts and massive exchange of
unimproved and improved germplasm throughout the world may not
always reflect the genetic diversity among parents of different geographic
origin. On the other hand, extensive hybridization in several national and
international breeding programs around the world has created vast
genetic diversity among the lines developed under given geographic
conditions and one can expect genetic diversity among parents from the
same geographic origin.
Biotechnological tools such as isozyme, RFLP and random amplified
polymorphic DNA (RAPD) are used to estimate genetic diversity among
rice cultivars (Chu, 1967; Shahi et ai., 1969; Pai et ah, 1975; Fu and Pai,
1979; Second, 1982; Glaszmann, 1987; Mackill 1995). Isozyme
polymorphism in parents and hybrids in relation to heterosis for
quantitative traits has been studied by a number of researchers. Xiao
(1981) Deng and Wang (1984); Yi et ah, (1984) have shown association of
heterosis with a certain esterase peroxidase pattern in FjS. However,
there was no association between the magnitude of heterosis in FjS and
genetic diversity among parents as determined by polymorphism at six
loci, viz. Est-9, Est-2, Amp-3, Sdh-1, Pgi-2 and Pgd-1 (Peng et ah, 1988).
Conclusive evidence of relationship between RFLP studies and genetic
diversity in rice is lacking.
Mahalanobis generalized distance (D^ statistics) based multivariate
analysis (Mahalanobis, 1936) is used to estimate genetic diversity by
classifying prospective parents of hybrids into various genetically diverse
clusters. Parental lines belonging to distantly located clusters are more
likely to give heterotic hybrids than parental lines belonging to the same
cluster. This technique has been used to classify rice cultivars by
Julfiquar et ah (1985), Vaidyanath and Reddy (1985), and Li and Ang
(1988), Li and Ang (1988) observed that panicle, grain number and
growth duration played a major role in estimating the genetic distance
among parents. Inclusion complex traits such as yield may bias the
estimation of genetic divergence (Julifquar et ah, 1985). Tropical
japónicas or javanica rices are japónicas adapted to tropical conditions
and genetically diverse from indica rices (Glaszmann, 1987; Khush and
Acquino, 1994). In a study conducted at IRRI hybrids derived from
indica/tropical japónica crosses were more heterotic than hybrids
derived from indica/indica and tropical japónica/tropical japónica
crosses (Table 2.7).
Virmani et ah (1991, 1994), on the other hand, observed a lower
degree of heterosis in indica/temperate japónica crosses compared to
crosses between indica/indica, implying that genetic diversity between

J.S, Nanda and S.S. Virmani 33
Table 2.7
Total biomass, grain yield, harvest index (HI), and 1000-grain weight
of intervarietal and intravarietal group hybrids and their inbreds
evaluated at IRRI, 1993 WS, and 1994 DS
Group Nuniber Total biomass
(g m'^)
Grains HI 1000-grain
wt (g)
1993 WS (spacing 20 cm x 10 cm)
TJ/I 3 1816a 890a 0.49a 31.0b
I/I 5 1540b 710b 0,46b 28.0c
TJ/TJ 5 1489b 643c 0,43c 32.6a
I 9 1418b 603c 0.42c 26.3d
TJ 8 1116c 412d 0.37d 28.8c
1994 DS (spacing 15 cm x 10 cm)
TJ/I 8 2047a 1030a 0.50a 28.1a
I/I 8 1834b 894b 0.48b 27.0b
TJ/TJ 8 1724bc 822c 0.48b 27.6b
I 8 1651c 726d 0,44c 24.4c
TJ 8 1453d 566e 0.39d 25.0c
Source: Virmani (1996),
TJ, tropical japónica; I, indica.
Data with suffix of same letter indicate that those data are not significantly different from
each other.
combinations. Hybrids between indica and japónica rices show variable
degrees of sterility. Ikehashi and Araki (1984) suggested the use of wide
compatibility gene(s) to overcome the hybrid sterility in indica/japónica
hybrids. Ikehashi and Araki (1984) showed that gamete abortion by an
allelic interaction at a locus (S5) caused hybrid sterility in S5i S5j but not
in S5n S5i or S5n S5j; S5i, S5j and S5n representing indica, japónica and
neutral allele respectively. Thus by incorporating the S5n allele into one
of the parents, sterility may be overcome. The wide compatibility locus,
S5n (WC) locus is closely linked with marker genes C (chromogen for
pigmentation) and wx (waxy endosperm) which are located on
chromosome 6 (Ikehashi and Araki, 1986, 1987). Subsequent to the
findings of Ikehashi and Araki (1984), a total of six loci, S-5, S-7, S-8, S-9,
S-15 and S-16, located on chromosomes 6, 4, 6, 7,12 and 1 respectively,
are known which can cause hybrid sterility in intervarietal hybrids
independent of each other and for which neutral alleles WC genes have
been identified in different rice cultivars (Yanagihara, et al., 1992; Wang
et al, 1993). Virmani (1996) has given a list of some WCVs identified
with the S locus in Japan, China and IRRI (Table 2.8).
Combining ability analysis has been useful in identifying potential
parents to produce heterotic hybrids. Generally speaking the probability
of getting heterotic hybrids is high in parental lines having high general
rnmViinmo- ahililv (GG A) flhidies Conducted at IRRI indicated that

34 Rice Breeding and Genetics: Research Priorities and Challenges
Table 2.8 Some WCVs identified in Japan, China and IRRI
WCV cultivar Varietal group WC gene Reference
Ketan Nangka TJ S5n Ikehashi and Araki (1986)
Calo toe TJ S5n Ikeháshi and Araki (1986)
CP SLO 17 TJ S5n Ikehashi and Araki (1986)
NK 4a J S5n Araki et al. (1988)
Norin PI 9a J S5n Arakai et al. (1990)
Banten TJ S7n Yanagihara et al. (1992)
N22 Aus S5n, S7n Yanagihara et al. (1992)
Dular Aus S5n, S7n Ikehashi and Araki (1987);
Vijayakumar and
Virmani (1993)
Padi Bujang Pendek TJ S5n Ikehashi and Araki (1987)
Aus 373 Aus S5n Ikehashi and Araki (1987)
DV 149 Aus . S5n Ikehashi and Araki (1987)
Kaladumani Aus S5n Ikehashi and Araki (1987)
DV 52 Aus S5n Ikehashi and Araki (1987)
AS 35 Aus S5n Ikehashi and Araki (1987)
Lepudumai Aus S5n Ikehashi and Araki (1987)
0 2428 J S5n Wang ef al. (1991)
TJ, Tropical japónica; J, japónica and aus: Rice variety grown is the aus rice growing season
in Bengal
GCA effects (Peng and Virmani, 1990), though sometimes heterotic
combinations were obtained from parents having low GCA effects
(Kumar and Saini, 1981). Srivastava and Seshu (1983) reported that occasionally
parents showing GCA did not result in heterotic combination.
Apparently the prediction of heterotic combinations of parents cannot be
made accurately based on combining ability studies alone.
Hybrid Rice Breeding Technology
Rice is a self-pollinated crop. Therefore, use of an effective male sterility
system to develop and produce hybrids in rice is imperative for the
success of hybrid rice breeding. Cytoplasmic-genetic male sterility
(CMS) system is extensively used in hybrid rice breeding. Recently,
concerted efforts have been made to develop an environmentally
sensitive genetic male sterility (EGMS) system and its application in
hybrid rice breeding. Apomixis, asexual seed production, a genetic tool
for developing true breeding hybrids with permanently fixed heterosis
has tremendous potential and needs sharp focus in the future research
strategy on hybrid rice.
Cytoplasmic-Genetic Male Sterility (CMS)

I
i
J.S. Nanda and S.S, Virmani 35
homozygotic recessive nuclear genes {rfrf) for fertility restoration and
sterility inducing factor (S) in cytoplasm makes plants male sterile. Thus,
only S rfrf individuals are male sterile. Possible cytoplasmic-genetic
constituents of plants and their pollen fertility behavior and accepted
designations are: S rfrf (sterile, CMS), N rfrf (fertile maintainer), S /N
RFRF ( fertile restorer), S Rfrf (fertile, hybrid). A CMS (A, line) is maintained
and multiplied by crossing with its corresponding mdintainer (B,
line). A cross of CMS and a restorer (R) line results in a commercial
hybrid.
Based on the genetic behavior, the CMS lines are classified into two
types, sporophytic and gametophytic. In the sporophytic male sterility
system, pollen sterility or fertility is determined by the genotype of the
sporophyte, and the genotype of the grains has no effect per se. When
tfie sporophyte genotype is S (rr), all the pollen grains will be abortive. If
the genotype is N (RR) or S (RR), all the pollen grains will be fertile. When
the sporophyte genotype is S (Rr), it produces two types of male gametes,
S(R), and S(r), and all the pollen grains are fertile. The CMS lines of WA
type and Gam type belong to the sporophytic type and have the following
characteristic features;
The F^ hybrid from the cross of A /R has normal pollen grains
with normal fertility. Segregation in fertility will occur in F2 and a
certain proportion of male sterile plants will appear in the
population.
• Abortion of pollen grains occurs at an earlier stage of microspore
development. Most of the grains look wrinkled and irregular,
• The anthers are milk-white in colour and water soaked and
indéhiscent,
• Male sterility is stable, but its restoring spectrum is relatively low.
• The first internode from the top of the culm is shorter, and the basal
part of the panicles is enclosed in the flag leaf sheath to a varying
degree.
In the gametophytic male sterility system, fertility of the pollen
grains is determined by the genotype of the gametophyte, viz, pollen.
The nuclear gene R or r in the pollen results either in fertility or sterility
respectively. The CMS lines, BT type, Tian 1 type, and Hong-Lien type
belong to the gametophytic type and have the following features:
• The Fi hybrid from the cross, A/R, has pollens of two genotypes,
R and r, in equal proportion,
• Pollen grains of genotype r are sterile.
• Pollen abortion occurs at the later stage of microspore
development.
TT-io O f a cio n r1 or* anH i-nHi>tiicrpnt

36 Rice Breeding and Genetics; Research Priorities and Challenges
• Its restoring spectrum is relatively broad.
• Panicles are not enclosed in the sheath.
Sampath and Mohanty reported for the first time in 1954 the role of
cytoplasm in causing male sterility in rice. However^ the first
cytoplasmic male sterile line used in the development of commercial
hybrids was developed in China in 1973 from a male sterile plant
occurring in a natural population of wild rice {Oryza sativa f spontanea),
(Yuan, 1977). The CMS plant was designated as wild rice with aborted
pollen (WA). Since then a number of CMS lines have been developed in
various program. Male sterility inducing cytoplasms were also
identified in various geographic forms of O. perennis (Rutger and
Shinjyo, 1980). The frequency of male sterile cytoplasm in Asian and
American strains was about 64% and 4% respectively, Li and Zhu (1988)
studied 300 strains of O. rufipogon and found 62 strains to possess male
sterility inducing cytoplasm. In China most of the cytoplasmic male
sterile lines have been primarily developed from the cytoplasmic source,
viz. O. sativa /. spontanea indica cultivars (Li and Zhu, 1988), although
some japónica rices, Ke Qing 3, Zhaotong-Beizigu, Ma-zoou-gu and O.
glaberrima Dan botus were also donors of sterility-inducing cytoplasm.
Yabuno (1977) identified three japónica cultivars; Akebono, Norin 19
and Omachi, to give cytoplasmic male sterile lines in combination with
two accessions of O. glaberrima. Pradhan et al. (1990) identified two new
CMS sources among indica cultivars, V20B and Sattari in crosses with
japónica rices. V20B is the maintainer of CMS-WA cytoplasm but a
source of sterile cytoplasm with japónica rice cultivar.
Among the several CMS lines evaluated until 1996 for their stability
of sterility and other desirable traits, only three lines, viz., IR58025A, IR
62829A and PMS 3A were found to be commercially usable on a large
scale. Even among these promising three CMS lines, IR 62829A has some
problems of stability of male sterility at high temperature and the outcrossing
in PMS 3A is observed to be low. Therefore, there is only one
reliable CMS line, IR 58025A, which has been used extensively for the
development of commercial rice hybrids. During the past two years some
other commercially usable CMS lines have been developed at IRRI.
Extensive use of and dependence on a unitary source of sterilityinducing
cytoplasm may lead to a sudden outbreak of pest and diseases
as in the case of corn. In pearl millet the evidence was only empirical and
not conclusive. Hence, serious efforts are made for the diversification of
the source of male sterility cytoplasm. Several interspecific crosses
involving cultivated O. sativa and closely related wild/weedy species of
A genome (O. rufipogon, O. nivara, O. barthii, O. glaberrima and O.
n rc .i-rT rv i-i-n ii4 -/i\ Ci.NQ,.>Í£»o r ^ o c o i s o o i - n r y o t o f i l i a r " i r f r » r a l C I Y I

J.S. Nanda and S.S. Virmani 37
the African species studied was found to possess sterile cytoplasm. In all
six male sterility sources were developed which contained the sterile
cytoplasm of either O. rufipogon or O. nivara (Table 2.9).
Table 2,9
Characteristics oi newly identified alternate CMS sources
Code Source Type of sterility Restorers Maintainers
RPMSl O. rufipogon Gametophytic Nil IR66,IR70, PMS 2B,V20 B
RPMS2 0. nivara Gametophytic Nil IR66, IR70
RPMS3 O. nivara Stained pollen IRBB7 PMS2E,IR62829B
(MS 577A-like)
RPMS4 0. nivara Sporophytic Nil IR66,PMS2B
RPMS5 0. nivara Sporophytic IRBB7 PMS2B, IR 62829 B
IR 66,1R70
RPMS6 0. nivara Sporophytic 1RBB7 PMS 2B,IR 62829 B
IR 66,IR 7
Source: Siddiq et. ui., 1996.
Of the six sources identified, three were found to be differ from the
WA system based on restoration response and pollen stainability.
Among these three, two are gametophytic types and one is sporophytic.
Of the remaining three sources, one is similar to MS 577A, while the
other two resemble the WA type with respect to restoration and
maintenance reaction.
During the past two decades about 20 CMS sources have been
identified. However, the WA source is predominantly used in the
production of commercial hybrids. Some of the recently identified
sources include CMS-ARC, O. perenniSf IR66707A (Dalmacio et ah, 1995),
O. ghimaepatitla, IR69700A ( Dalmacio et al., 1996) and gamma ray
induced mutant from IR62829 B (IRRI, 1995). Pradhan et al. (1990a)
identified two new sources, V20B and Sattari, through indica/japónica
hybridization. V20B is a maintainer of CMS-WA source but itself was
found to be a source of CMS with japónica cultivar Zhunghua-1 (Pradhan
et al, 1990b). More recently, two CMS sources from O. rufipogon and
one from O. nivara were identified. These are designated as RPMS-1
(0. rufipogon), RPMS-2 (O. nivara) and RPMS-4 (O. rufipogon). Six CMS
lines have been developed from these sources (Table 2.10). These CMS
lines are highly stable except RPMS-2 and panicle exertion is almost
complete, unlike the WA source. However, no restorers have been
identified from the cultivated germplasm for these lines.
Two CMS lines, Pushpa A and Mangla A, are available with MS577A
cytoplasm derived from O. rufipogon. However, these lines do not
possess good agronomic traits and floral characteristics.
With the intensification of research on hybrid rice, a number of

have been classified based on genetic behavior, relation between
restorers and maintainers, and the morphology of sterile pollen grains
(Yuan and Fu, 1996).
According to the relationship of the restorer and maintainer lines,
the CMS lines may be classified into three groups. Group I is
representable by WA type, group II by Hong-Lien type and group III by
the BT type. The frequency of obtaining maintainer lines for the WA
type is high in early maturing dwarf indica varieties cultivated in China,
but low in IRRI lines. However, the frequency of obtaining restorer lines
is high in IRRI lines and varieties originating from Southeast Asia and
southern part of China.
CMS lines of the Hong-Lien type are developed by backcrossing
red-awned wild rice as the female parent with the recurrent male parent
Lian-Tang-Zao. The relationship between restorers and maintainers of
the Hong-Lien type is roughly contrary to that observed in the WA type.
Some early maturing dwarf indica varieties such as Zhen-Shan 97, ER-
Jiu-Ai and Xian-Feng 1, which are maintainers for WA CMS lines, are
restorers for Hong-Lien CMS lines. On the other hand, some restorers of
WA CMS lines such as Peta, Tai-Ying 1, Indonesia 6 and Xue-Gu-Zao,
are good maintainers of Hong-Lien CMS lines. Many IRRI lines, such as
IR24, IR26 etc. partially restore fertility in the Hong-Lien type.
The relation between restorers and maintainers for the CMS lines
developed by using the cytoplasm of Tian-Ji-Du, e.g. Tian-Ai A, is
similar to that of the Hong-Lien type.
The CMS lines Taichung 65, some japónica CMS lines such as Li-
Ming and Feng-Jin with sterile cytoplasm transferred from BT and CMS
lines of japónica rice Tian 1, belong to this type. Most of the japónica
varieties from Japan and China are good maintainers for the BT type
CMS line. The restorer eenes in the existine restorer lines of iaoonica rice
38 Rice Breeding and Genetics: Research Priorities and Challenges
Table 2.10
Desirable features of the new CMS lines possessing alternate
sources of male sterility
CMS line
Stigma exsertion
(%)
Panicle exsertion
(%)
Outcrossing
(%)
Male sterility
(%)
RPMS 1-1 40.56 99.5 35.5 S
RPMSl-2 38.60 96.5 28.5 S
RPMSl-3 42.50 98.6 30.5 s
RPMSl-4 46.46 99.5 35.5 s
RPMS 2 42.40 98.6 34.2 US
RPMS 4 39.46 98.9 31.6 s
S, Stable; US, unstable.
Source', Siddiqei ah, 1996.

To develop rice hybrids, it is necessary to have effective restorer lines.
J.S. Nanda and S.S. Virmani 39
from sporogenous cells results in male sterility of the pollen-free type/
whereas failure in the mechanism during microspore development and
pollen maturation results in various types of pollen abortion.
Classification of CMS lines based on the morphology of pollen grains
stained with I-KI solution falls into the following categories:
• Typical abortion type: The pollen grains are irregular in shape
and not stainable with I-KI solution. Pollen abortion occurs
relatively early^ mainly at the single nucleus stage. It is called the
uninucleate abortion type. Most CMS lines of the WA type belong
to this type.
• Spherical abortion type: The pollen grains are spherical and not
stainable with I-KI solution. Pollen abortion occurs approximately
at the two-nuclei stage. This type is called the binucleate abortion
type. CMS lines of the Hong-Lien type belong to this category.
• Stained abortion type: Pollen grains are spherical^ but much
smaller compared to the normal pollen^ partially or lightly stained
with I-Kl solution. Pollen abortion takes place at the three-nuclei
stage. This type is called die trinucleate type. CMS lines of the BT
type belong to this category.
Virmani and Shinjyo (1988) proposed an interim designation for the
various cytoplasmic sources known to induce male sterility. The CMS
sources are designated in principle according to the name of the cultivar
from which the cytoplasmic factor inducing male sterility is derived
(Table 2.11).
Virmani and Shinjyo (1988) also proposed a model for the
identification of genetic differences among cytoplasm and restoring
genes, which involves the development of CMS lines possessing
different cytoplasm and their restorer lines in an isogenic genetic
background by recurrent backcrossing. These lines are then intercrossed
and their Fj plants are evaluated for pollen and spikelet fertility. If the Fj
shows differential reaction to different Ri genes, the cytoplasm is
considered different. To determine the allelic relationship among
different Rf genes, two different restorer lines (cms-A Rfl Rfl and cms-B
Rf2 Rf2) are intercrossed; their Fj (cms-Afl Rf2) is pollinated by the
maintainer having no restorer genes (N-rfl rfl and/or N-rf2 rf2) and the
backcross progeny is evaluated for pollen fertility. The segregation
pattern of the backcross progeny would indicate whether the Rf genes
are allelic or non-allelic.
Fertility Restoration

The frequency is also higher among indica rices compared to japónicas.
Li and Zhu (1988) observed that among the three ecotypic rice cultivarse
40 Rice Breeding and Genetics: Research Priorities and Challenges
Table 2.11 Male-sterile cytoplasm, its sources and designation
Cytoplasm Strain Nuclear donor Reference Interim
donor
designation
(1) (2) (3) (4) (5)
0. sativa Chinsurah boro II Taichung 65 Shinjyo and Omura, 1966a cms-bo
0. sativa Lead rice Fujiska 5 Watnabe et al., 1968 cms-ld
0. sativa Tadukan NorinS Kitamura, 1962a cms-TA
O. sativa Chinese strain Fujusaka Katsuo and
f. spontanea Mizushima, 1958 cmsCW
0. sativa Wild abortive Several Lin andYuan, 1980 cmsWA
f. spontanea
0. sativa Red-awned wild Lien- Lin and Yuan, 1980 cmsHL
f, spontanea Tong-Tsao
O. sativa Akebono 0 . glaberrima Yabimo, 1977 . cms-ak
O, rufipogon W 1080 (India) Taichung 65 Shinjyo ef«/., 1981 cmsWlS
0. rufipogon W 1090 (India) Taichung 65 Shinjyo and Matomura, cmsW19
1981
0, rufipogon KR7 Taichung 65 Cheng and Huang, 1979 cmsKR
0. sativa
f. spontanea YaChe Guang Xuan 3 Virmani and Wan, 1988 cmsYC
0. sativa Tian Dong Zhen Shan 97 Virmani and Wan, 1988 cmsTD
0. sativa Lie Zhou Zhen Shan 97 Virmani and Wan, 1988 cmsLZ
0. sativa Indian Jin Nan Te 43 Virmani and Wan, 1988 cmsIN
0, sativa Dong Pu JinNan Te43 Virmani and Wan, 1988 cmsDP
O. sativa JunNiya Chao Yang 1 Virmani and Wan, 1988 cmsJNY
0. sativa HePu Li Ming Virmani and Wan, 1988 cmsHP
0. sativa Teng Qiao Er-Jing-Qing Virmani and Wan, 1988 cmsTQ
0. sativa San Ya Jing Yin 1 Virmani and Wan, 1988 cmsSY
0. sativa Rao Ping 6964 Virmani and Wan, 1988 cmsRP
0. sativa Guangzhou 6964 Virmani and Wan, 1988 cmsGZ
0, sativa Dwarf aborted Xue Qin Zhao Virmani and Wan, 1988 cmsDA
0, sativa Taichung Nativel Pankhari 203 Athwal and Virmani, 1972 cmsTN
O. sativa Gamibiaca Chao Yang 1 etc. Lin and Yuan, 1980 msGAM
0. sativa BircofPI 279120) Calrose Erickson, 1969 msBI
0. sativa ARC 13728-16 IR101179-2-3-1 IRRI, 1986 cmsARC
O. sativa Eshan Ta Bei Cu Hong Mao Virmani and Wan, 1988 cmsSTB
Ying
0. sativa Tian Ji Du Fujisawa 5 Virmani and Wan, 1988 cmsTJD
0, sativa IR24 Xiu Ling Virmani and Wan, 1988 cmsIR24
0. sativa Jing Chuan Nao NanTai Geng Virmani and Wan, 1988 cmsJCN
0. sativa ShengQi Nong Ken 8 Virmani and Wan, 1988 cmsSQ
0> sativa Li Up Jing Yin 83 Virmani and Wan, 1988 cmsLU
0. sativa Zhao Jin Feng Lan Bery Virmani and Wan, 1988 cmsZJF
0. sativa Zhao Tong BeiKe Ching 3 Virmani and Wan, 1988 cihsZTB
0. sativa Dissi Hatif Zhen Shan 97 Virmani and Wan, 1988 cmsDIS

J. s. Nanda and S.S. Virmani 41
cultivars. Bulu and Tjerehh cultivated in Java, Indonesia, Tjereh
cultivars had a higher frequency of restorers. Bulu rices while showed
weak restoration. In Asia, restorers were found mainly in South and
Southeast Asian countries and in southern China.
The genetics of fertility restoration has been studied by several
workers and was recently reviewed by Virmani (1996). The effect of
restorer gene(s) for CMS-bo and CMS-D cytoplasm was gametophytic
causing partial pollen fertility, but normal spikelet fertility in Fj hybrids.
On the other hand, the effect of restorer gene for CMS-WA cytoplasm was
sporophytic which gave normal pollen and spikelet fertility in Fj
hybrids. The inheritance of fertility restoration in CMS-bo and CMS-D
cytoplasm was monogenic dominant and the two genes were allelic (Hu
and Li, 1985). Teng and Shen (1994) reported that the fertility restoration
for CMS-bo was controlled by a dominant gene RF-1 carried in the
restorer line C57, or by an incompletely dominant gene in Z H 157 which
was diffefent from RF-1. Wang (1980) reported a single dominant gene
restoring fertility in cytosterile Zhen Shan 97A, whereas all other studies
indicated that the fertility restoration was controlled by two dominant
genes (Govindraj and Virmani, 1988; Teng and Shen, 1994) with
sporophytic effect and that one of the genes had a stronger effect than the
other. Li (1985) and Li and Yuan (1986) reported that of the two restorer
genes of IR24, one gene, RlR l, might have been inherited from a late
indica variety from China, while the other, R2R2, from SLO 17. The two
genes showed differential effects in their strength for restoration ability.
The allelic test of restorer genes, present in six restorer varieties; (IR26,
IR36, IR54, IR9761-19-1, IR2707-105-2-2-3 and IR42) indicated the
presence of four groups of restorers possessing different pairs of restorer
genes (Govindraj and Virmani, 1988). These authors suggested that the
probable sources of R genes in two popular varieties, IR36 and IR42 are
Cina, Latisail, Tadukan, TN1, TKM 6, PTB18, and SLO 17. The existence
of a large number of restorer ¡genes explains the high frequency of R lines
among the elite indica breeding lines for the CMS-WA cytosterile system.
Shinjyo (1975) located the Rf gene fcir CMS-bo cytoplasm on
chromosome C, presently designated as chromosome 10 using trisomie
analysis. Yoshimura et al. (1982) located the Rf gene for CMS-bo
cytoplasm on chromosome 7 using the translocation inethod. Shinjyo and
Sato (1994) also located the restorer gene (Rf-2) for CMS-L cytoplasm on
chromosome 2 using primary trisomie and linkage tester lines. Bharaj et
al. (1995) using trisomie analysis located the stronger restorer gene (Rf-
WA-1) for CMS-WA cytoplasm on chromosome 7, while the weaker
gene (Rf-WA~2) was located on chromosome 10. Shinjyo (1975) reported

42 Rice Breeding and Genetics: Research Priorities and Challenges
G enetic M ale Sterility
The cytoplasmic-genetic male sterility system is extensively used in the
production of commercial rice hybrids. It is a three-line system involving
a CMS source, a maintainer and a restorer. It is an effective system.
However, there are some drawbacks, such as the sources of developing
new CMS lines are rather limited and poor. Presently the WA type of
CMS system is extensively used. The unitary cytosterility system
associated with the narrow genetic base poses latent danger of
susceptibility to dangerous diseases and insect pests. In the long run this
may bring about serious crop losses. Moreover, the sterility of japónica
CMS lines (BT type) now used to produce japónica rice hybrids is not
stable enough to produce pure seeds.
Shi (1981) reported that Shi Ming Song found a male sterile plant in
the field of a japónica rice cultivar Nong-ken 58 in Hubei province, China.
It behaved is a male sterile when plants headed under long daylength,
but a male fertile under short daylength. The degree of male sterility
was 99-100% at heading under artificial light of more than 14 h, but
plants were male fertile when grown under artificial light of less than 13
h 45 min. (Lu and Wang, 1988). It was called Hubei-photoperiodsensitive
genetic male sterile rice. Yang et al (1989) induced *a
photoperiod-sensitive genetic male sterile (PGMS) line (5460) in indica
rice IR54. Subsequehtly, temperature-sensitive genetic male sterility
(TGMS) was also reported (Zhou et al., 1988, 1991; Wu et at, 1991;
Virmani and Voc, 1991).
Several PGMS and TGMS lines have been developed in China,
Japan, the USA and by IRRI (Table 2.12). Satynarayana et al (1995)
reported a new source of the TGMS line in India among spontaneously
occurring sterile plants from indica cultivar lET 10726. Orad and Hu
(1995) used ethylmethane sulfonate to induce PGMS in rice cultivar
M201. PGMS lines Nongken 58S and X-88 are completely pollen under
long day (day length below 13.75 h). TGMS lines show complete pollen
sterility when the maximum temperature during the day is above 27-
29°C and partially fertile when the maximum day temperature is lower
than 27“C. However, the critical temperature has been found to vary
depending on the source of the TGMS gene (Lu et al, 1994).
Both PGMS and TGMS are referred to as environment sensitive
male sterile (EGMS) mutants. Wang et al (1991a, b) Maruyama et al,
(1991), Borkakati and Virmani (1989) reported both TGMS and PGMS to
be monogervic recessive traits. Sun et al (1989) and Maruyama et al.
(1991) designated the TGMS genes as tmsl and iws2, respectively.
Borkakati and. Virmani (1996) reported that the two TGMS mutants,

The TGMS gene in Norin PL12 has been designated as tms2 (Kinoshita
1992). The TGMS gene in the mutant IR32364TGMS is designated as
fms3.
The PGMS and TGMS lines are currently used to develop two-line
rice hybrids. This system gives higher frequency of heterotic hybrids
compared to the CMS system because of limited restrictions in the choice
of parents. In China, about 11 usable S lines (2 japónica PGMS and 9
indica TGMS) have been developed and registered and 5 two-line hybrid
rice varieties have been released for cultivation. The area under two-line
hybrid rice varieties in China two increased from 60,000 ha in 1994 to
about 330,000 ha in 1996 (Tran and Nguyen, 1998), The two-line hybrid
varieties have a yield advantage of 5-10 % higher than the commercially
cultivated three-line hybrid varieties.
The PGMS system is useful in temperate regions where the
daylength differences during the rice-growing season are striking. The
TGMS system, on the other hand, is useful under tropical conditions,
where the daylength differences are marginal, while the temperature
differences between high and low altitude are considerable.
€
J.S. Nanda and S.S. Virmani 43
Table 2.12
Some of the PGMS and TGMS lines developed in China,
Japan, the USA and IRRI
Une
Varietal
group
Developed
by
Developed
at
Type
Fertility induction
conditions
NongkenSSS J Spontaneous
mutation
China EGMS Daylength lovi^er than
13.75 h
Anong S I Spontaneous China TGMS Temperature 27'*C
mutation
Hennong S I Crossbreeding
China TGMS Temperature
below 29“C
5460 S I Irradiation China TGMS Temperature
below 29“C
R59T S
I Irradiation China TGMS Temperature
below 29“C
IR32364-20-1-3-2B
I Irradiation IRRI TGMS Temperature
below 29“C
Norin PL 12 J Irradiation Japan TGMS Temperature
below 28“C
IVA I Crossbreeding
China TGMS Temperature
below 24“C
Dianxin lA J CMS China TGMS Temperature
below 22®C
EGMS J USA PGMS Daylength

44 Rice Breeding and Genetics: Research Priorities and Challenges
APOMIXIS
Apomixis is an asexual method of reproduction in which the embryo
(seed) develops without the union of male and female gametes. It
bypasses meiosis and syngamy in the female gametophyte to produce
embryos genetically identical to the maternal parent. Apomixis is the
genetic tool to develop true breeding rice hybrids. The application of
apomixis in rice would enable farmers to use the harvest of a hybrid crop
as seed for the subsequent hybrid crop without experiencing genetic
segregation and inbreeding depression. Therefore, it will enable even the
resource-poor farmers to benefit from hybrid rice.
The indicators of apomixis are:
• Identical maternal progeny from plants of cross-pollinated species
or progeny of Fi crosses.
• Limited or no genetic variation in the F2 population of a cross
between two distinct parents.
• Recessive genotypes from a cross of parents with recessive genes
pollinated with a parent possessing a dominant marker gene.
• Unusual high seed fertility in aneuploids, triploids and wide
crosses normally expected to be sterile.
• Aneuploid chromosome number or structural heterozygosity
remaining consistent from parents to progeny.
• Multiple seedlings per seed, multiple stigmas, multiple ovules per
floret and doubled or fused ovaries.
There are reports of mutants in rice with twin seedlings per seed
(Yuan et al, 1990; Sharma and Virmani, 1990) and multiple pistillate
ovaries (Suh, 1985,1988). Occurrence of apomixis in interspecific crosses
in rice has been reported from China (Chen et al., 1988), But, the
authenticity of this report has yet to be confirmed.
The potential of apomixis in the commercial exploitation of heterosis
in rice bears great promise, Khush and Gquino (1994) have suggested
three strategies to intensify the search for apomixis in rice. These include:
• Analysis of the tetraploid wild germplasm of Oryza.
• Induction of mutations for apomixis.
• Use of molecular approaches to engineer apomixis.
BIOTECHNOLOGICAL APPLICATIONS
Recent advances in biotechnology have opened new avenues in hybrid
rice breeding. Brar et al. (1994) have outlined several applications of
ciirh afi anther culture, embrvo rescue, protoplast fusion.

Table 2.13
Application of biotechnology in hybrid rice'
Technique
Anther culture
Embryo rescue
Protoplast fusion line
o p ia tic einbryogenesis
Molecular markers
Genetic transformation
Applications
Extraction of high-yielding inbred lines from 'superior
hybrids, purification of male sterile, maintainer arid restorer
lines.
'
Overcoming incompatibility to produce hybrids between
wild and cultivated species, Deriving back-cross progenies to
develop alloplasmic lines for diversification of CMS sources,
Transfer of genes for apomixis from wild species into elite
breeding lines.
Expeditious transfer of CMS into elite breeding, development
of cybrids between otherwise sexually incompatible species.
Production of artificial seeds for mass propagation of true
breeding hybrid varieties.
Tagging genes for wide compatibility, fertility restoratiôn,
thermosensitive male sterility, apomixis, and identifying
QTLs for heterosis to facilitate marker-based selection;
choosing parents based on RPLP diversity to obtain highly
heterotic combinations.
Transfer of cloned genes governing apomixis for producing
true breeding commercial hybrids; exploitation of genetically
engineered nuclear male sterility and fertility restoration
systems to produce hybrid varieties.
FUTURE OUTLOOK
The development of semidwarf varieties of rice in the 1960s led to
significant increases in the yield of rice. The yield potential of semidwarf
high-yielding varieties in the tropics is 10 tha'^ during the dry season
and 6.5 tha'^ during the wet season. The maximum yield potential was
estimated to be 9.5 tha'^ during the wet season and 15.9 tha'^ during the
dry season (Yoshida, 1981). Since the development of semidwarf
varieties, however, there has been little improvement in the yield
potential of rice. Major efforts in the past three decades have been
towards the incorporation of disease and insect resistance, shortening of
growth duration and improvement in grain quality (Khush, 1994).
Exploitation of heterosis through hybrid rice breeding has offered a
potential venue for increasing rice yield. China has made remarkable
progress in the exploitation of hybrid rice-breeding technology to boost
rice production.
Realization of the potential of hybrid rice technology is greater in
countries with a higher proportion of irrigated rice area, a high land:
labor ratio and under transplanting conditions where the seed

46 Rice Breeding and Genetics: Research Priorities and Challenges
and where rice is direct seeded if, yields of hybrid seed were around 3
tons and above per hectare and the seed cost reduced.
The current level of heterosis in indica rice hybrids is around 15-
20%. However, a higher level of heterosis is desirable. The magnitude of
heterosis depends upon the genetic diversity between the two parents of
the hybrid. In the last three decades, the genetic diversity among the
improved indica rices has narrowed because of the massive international
exchange of germplasm. However, there has been little gene flow
between indica and japónica rices and these two races of rice have
remained distinct. A higher level of heterosis has been observed in
hybrids between indica and tropical japónica for yield than indicaindica
hybrids. However, this would require tackling associated
problems such as intervarietal hybrid sterility, lodging, late maturity
and grain quality. Deployment of wide compatibility genes in breeding
indica/japónica hybrid is essential. Tagging of WC genes with molecular
markers to assist in marker-aided selection for this trait would be very
useful. Rice hybrids with the desired disease and insect resistance and
acceptable grain quality can be developed by an appropriate choice of
parental lines.
The CMS system has been successfully used in the development of
rice hybrids. Though several CMS sources have been identified, only a
few have been commercially exploited. It is very essential to genetically
diversify the CMS sources to avert possible incidences of disease and
insect epidemics. There is no dearth of restorers among the elite indica
rice cultivars. However, the restorers among the japónica rices are scarce.
It would be useful for the japónica rice hybrid program to find a new
CMS source for which sufficient restorers are available. The efficiency of
breeding of maintainer and restorer lines in indica rice can be accelerated
by developing genetically diverse maintainer and restorer populations
using male sterility-facilitated recurrent selection (Virmani et al, 1994).
Protoplast culture in ricé makes it possible to produce cybrids which
enable immediate transfer of cytoplasmic male sterility into elite rice
cultivars (Akagi and FujimUra, 1994; Brar ef al, 1994). This approach
needs to be explored to develop genetically diverse CMS lines
expeditiously.
The varieties developed during the past few decades have made a
significant impact in the favorable environment. Significant heterosis
observed for vegetative vigor and root characteristics suggests that
hybrid rice technology may have a potential in abiotic stress
environments (rainfed lowland, flood-prone, drought-prone, low

J.S. Nanda and S.S. Virmani
rice. Hybrid seed yields up to 5.8 tha'^ have been reported in China and
up to 2.5 tha'^ have been reported in other countries. Strategies such as
genetic improvement for flowering behavior of seed and pollen parents
(selection seed parents with long, exserted stigma, longer duration and
wider angle of floret opening, small and horizontal leaves, selection of
pollen parents with high percentage of residual pollen per anther after
exsertion, high pollen shedding potential through increasing number of
blooming spikelets per unit area) have to be vigorously perused.
Incorporation of elongated uppermost internode gene in male sterile
lines would be useful in improving panicle exsertion which would
eliminate the application of GA3 and reduce the cost of seed production.
The success story of hybrid rice in China has established beyond
doubt that hybrid rice technology has the potential to boost rice
production. However, the major constraints for its adoption are:
• need to buy fresh hybrid seed every planting season;
• high cost of hybrid seed;
• need to establish seed production infrastructure in developing
countries.
Farmers will buy hybrid seed at a price higher than that of inbred
seed only if there is a cost: benefit ratio of 1 ; 4 (Khush, 1994). This will
encourage investment by the government, private and cooperative
organizations in the seed industry.
Discoloration of hybrid seeds is a major problem. In China and
northwestern India, CMS lines in hybrid seed production plots have been
found to have a higher incidence of seed-borne diseases such as paddy
bunt and false smut compared to pollen parents. This not only
deteriorates the quality of the seed, but can also cause serious outbreak of
the disease in commercial crops of hybrid rice and therefore needs serious
attention.
Apomixis is the ultimate genetic tool to develop true breeding
hybrids and facilitate commercial exploitation of heterosis. Through
apomixis, heterosis can be permanently fixed. Farmers need not purchase
hybrid seed every year provided apomixis is successfully incorporated in
the hybrids. Research efforts in the exploration of apomixis and its
incorporation in the hybrid rice development technology need to be
intensified.
References
Akagi, H. and Fujimora, T. 1994. CMS Transfer in japónica varieties with cybrid method In:
Hybrid Rice Technolosiy: New Developments,and Future Prospects S.S. Virmani (ed.) IRRI,

48 Rice Breeding and Genetics: Research Priorities and Challenges
Akita'/ S. 1988. Physiological basis of heterosis in rice. In; '‘Hybrid Rice/' pp. 67-77. IRRI,
Manila, Philippines.
Araki, H., Ikehashi, H., Toya, K. and Matsumoto, S. 1990. Development of wide compatibility
rice line Norin PL 9. }pn. Agrie. Res. Q, 24: 78-81.
Araki, H., Toya, K. and Ikehashi, H. 1988. Role of wide compatibility genes in hybrid rice
breeding. In; Hybrid Rice, pp. 79-83. IRRI, Manila, Philippines.
Athwal, D. S. and Virmani, S. S. 1972. Cytoplasmic male sterility and hybrid breeding in rice.
In: Rice Breeding, IRRI, Manila, Philippines pp. 615-620.
Berger, E. 1976. Heterosis and the maintenance of enzyme polymorphism. Amer. Nat.llO:
823-839.
Bharaj, T. S., Virmani, S.S. and Khush, G.S. 1995. Chromosomal location of fertility restoring
genes for 'Wild Abortive' cytoplasmic male sterility using primary trisomics in rice.
Euphytica 83; 169-173.
Borkakati, R. and Virmani, S.S. 1993. Inheritance of a thermo-sensitive genic male sterile
mutant of indica rice. Rice Gene Newslett., 10:92-94.
Brar, D.S., Pujimura, T„ McCouch, S. and Zapata, F.J. 1994, Application of biotechnology in
hybrid rice. In '.Hybrid Rice Technology; New Developmental^ Future Prospects S.S. Virmani,
(ed.), IRRI, Manila, Philippines pp. 51-62.
Brar, D. S., Guo, Z. Y., Ahmed, M. I., Jachuck, P. J. and Virmani, S. S.1996. Diversification of
CMS system to improve sustainability of hybrid rice technology. Paper presented 3rd
Inti. Symp, Hybrid Rice, Indian Council of Agricultural Research, November 14-16,1996.
Chauhan, J.S., Virmani, S. S., Aquino, R. C. and Vergara, B. S. 1983. Evaluation of hybrid rice
for ratooning ability. Inti; Rice Res. Newslett. 8(6): 6.
Chen, J. S., Lin, Y. and Jun, C. W. 1988. A study of apomixis in rice. Genet. Manipulations Crops
Newslett. 4(2); 3~10.
Chen, Yinhui, Cai, Jumai and Lu Haoran. 1987. Effects of male sterile cytoplasmic interaction
on the genetic performance of rice. /. Fujian Agrie. Coll.16 (3): 190-197.
Cheng, Y. K. and Huan, C.S. 1979. Studies on cytoplasmic-genetic male sterility in cultivated
rice (Oryzfl sativa L.). I. Effect of different cytoplasm sources on male abnormalities at
anthesis. /. Agrie, Assoc. China 106:11-22.
Chopra, K. 1996. Global assessment of hybrid rice technologies for breeding and seed
production, Proc. 18th session Inti. Rice Comm., Rome, Italy, 5-9 September 1994, pp. 23-
36.
Chu, Y.E. 1967. Variation in peroxide isozymes of On/za perennis and O. sativa. }pn. i. Genet.
42: 233-240,
Comstock/ R, E, and Robinson/ H.F. 1952. Estimation of average dominance of genes. In;
Heterosis J.W. Gowen fed,)/ Iowa State College Press, Ames, lA pp. 494-^516.
Dalmacip, R. D,, Brar, D, S., Virmani, S.S, and Khush, G. S. 1996. Male sterile line in rice O.
sativa L , developed from O, glumaepatula cytoplasm. Inti Rice Res. Newslett. 21(1): 22-23.
Dalmado, R.D., Brar, D.S., Ishii, T., Sitch, L.A., Virmani, S. S. and Khush, G. S. 1995.
Identification and transfer of new cytoplasmic male sterility source from O. perennis into
indica rice (Oryza sativa 1.). Euphytica 82:221-225.
Deng, H. and Wang, G. 1984. A study on prediction of heterosis in crops, II. Analysis of
heterosis of rice and its esterase zymogram patterns, complementary patterns and
artificial hybrid zymogram patterns. Huan Agrie. Set. 3; 1-5. (in Chinese).
Erickson, J.R. 1969. Cytoplasmic male sterility in rice {Oryza sativa L). Agron. Abs. B.
Fu, Q. 1985. Studiés on the inheritance of restoring genes for wild abortive cytoplasmic
sterility in rice fOruzflfiiJfmfl-suBfi ho ela« >■ • - ...........

J.S. Nända and S.S. Virmani
Pu, P.Y. and Pai, C. 1979. Genetic studies on .isozymes in rice plant 2. Classification and
geographic distribution of cultivated rice through isozyme studies,/. Agrie. Assoc. China
107:1"16 (in Chinese with English summary).
Gallis, A. 1988. Heterosis: its genetic basis and its utilization in plant breeding. Euphytica
39(3); 95-104.
Glaszmann, J.C. 1987. Isozymes and classification of Asian cultivated rice varieties. Theor.
Appl. Genet.74:: 21-30.
Govinda Raj, K. and Virmani, S.S. 1988. Genetics of fertility restoration of WA type
cytoplasmic male sterility in rice. Crop. Sei, 28; 7&7~792.
Hu, J. and Li, Z. 1985. A preliminary study on the inheritance of male sterility of rice male
sterile lines with four different kinds of cytoplasms. /. Huai^oung Agrie. Coll. 4(2): 15-22.
Ikehashi, H. and Araki, H. 1984. Varietal screening of compatibility types revealed in Fj
fertility of distant crosses in rice. Jpn. J, Breed. 34:304-313-
Ikehashi, H. and Araki, H. 1986. Genetics of Fj sterility in remote crosses of rice. In: Rice
Genetics, IRRI, Manila, Philippines pp. 119-130,
Ikehashi, H. and Araki, H. 1987. Screening and genetic analysis of wide compatibility in
hybrids of distant crosses in rice, Oryza sativa. Tech. Bull. 22. Tropical Agrie, Res. Center,
Japan.
IRRI. 1986. Annual Report for 1985. IRRI, Manila, Philippines.
IRRI. 1995. Program Report for 1994. IRRI, Manila, Philippines.
Jinks, J.L. 1983. Biometrical genetics of heterosis. In: "Heterosis—Reappraisal of Theory and
Practice" R. Frankel, (ed.). Springer-Verlag, Berlin, Heidelberg, New York, pp. 1-46.
Jones, J.W. 1926. Hybrid vigor in rice. J. Amer, Soc. Agron.l8:424-428.
Julfiquar, A. W., Virmani, S. S. and Carpena, A. L. 1985. Genetic divergence among some
maintainer and restorer lines in relation to hybrid breeding in rice (Oryza sativa L ) Theor.
Appl. Genet. 70:671-678.
Katsuo, K. and Mizushima, U. 1958. Studies on the cytoplasmic difference among rice
varieties, Oryza sativa L. On the fertility of hybrids obtained reciprocally between
cultivated and'wild varieties. Jpn. f. Breed. 8:1-5 (in Japanese).
Khush, G.S. 1994. Increasing the genetic yield potential of rice; prospects and approaches.
Inti. Rice Comm. Neioslett, 43:1-8, FAO, Rome, 1995.
Khush, G.S. and Aquino, R.C. 1994. Breeding tropical japónicas for hybrid rice production.
In: Hybrid Rice: New Developments and Future Prospects S.S.Virmani (ed.), IRRI, Manila,
Philippines pp. 1-21,
Kim, C. H. 1985. Studies on heterosis in Fj rice hybrids using cytoplasmic-genetic male
sterile lines of rice (Oryza sativa L.), Res. Rep. Rural Dev. Administration, Suxoeon, Korea
27(1): 1-33.
Kinoshita, T. 1992. Report of the committee on gene symbolization, nomenclature and
linkage groups. Rice Genet. Neufslett. 9:2-4.
Kitamura, E. 1962, Studies on cytoplasmic sterility of hybrids in distantly related varieties of
rice, Oryza sativa L. I. Fertility of Fi hybrids between strains derived from a certain
Philippine x Japanese variety crosses and Japanese varieties. Jpn. J. Breed. 12; 81-84.
Kumar, I, and Saini, S.S. 1981. Diallel analysis in rice combining ability for various
quantitative characters. Genet. Agrie. 35: 243-252.
Li, Y.C. 1985. The pedigree analysis of the inheritance of restoring genes in IR24. Sei. Agrie.
Sin 1:24-31.
T.i. Y.C. and Ane. S. 1988. Genetic distance and heterosis in japónica rice. In; Hybrid Rice,

50 Rice Breeding and Genetics: Research Priorities and Challenges
Li, Z. and Zhu, Y. 1988 Rice male sterile cytoplasm and fertility restoration. In: Hybrid Rice,
IRRI, Manila, Philippines pp. 85--102.
Lin, S.C- and Yuan, L.P. 1980, Hybrid rice breeding in China. In: Innovative Approadies to Rice
breeding, IRRI, Manila, Philippines pp. 35-51.
Lu, X. and Wang, J. 1988. Fertility transformation and genetic behavior of Hubei
photoperiod'Sensitive genetic male sterile rice. In: Hybrid Rice. Proc. Int. Symp. Hybrid
Rice. Changsha, Hunan, China, Oct 6-10,1986. IRRI, Manila, Philippines, pp 129-138.
Lu, X.G., Zhang, G., Maruyama, K. and Virmani, S.S. 1994. Current status of two-line method
of hybrid rice breeding. In: Hybrid Rice Technology: New Developments and Future Prospects,
S.S. Virmani (ed.), IRRI, Manila, Philippines, pp. 37-50.
Mackill, D.J, 1995. Classifying japónica rice cultivars with RAPD. Crop. Sci.35:889-894.
Mahalanobis, P.C. 1936, On the generalized distance in statistics. Proc. Natl. Inst. Sci. India B.
2: 49-55.
Maruyama, S., Mijazato, K. and Nose, A. 1985. Basic studies in utilization of hybrid vigor in
rice. V. Reciprocal difference and heterosis in hybrid seeds and plants./p«. /. Crop. Sci.
54(4): 318-323.
Maruyama, K., Araki, H, and Kato, H. 1991. Thermosensitive genetic male sterility induced
by irradiation. In: Rice Genetics II, IRRI, Manila, Philippines, pp. 227-232,
Orad, J, H. and Hu, J. G. 1995. Inheritance and characterization of pollen fertility in
photoperiodically sensitive rice mutants. Euphytica 82:17-23.
Pai, C., Endo, T. and Oka, H. 1.1975, Genic analysis for acid phosphatase isozymes in Oryza
perennis and O, sativa. Can. /. Genet. Cytol. 17:637-650. .
Peng, J. Y, and Virmani, S. S. 1990. Combirung ability for yield and four related traits in
relation to hybrid breeding in rice, Oryza 27:1-10.
Peng, J.Y., Glazmann., J.C. and Virmani, S.S. 1988. Heterosis and isozyme divergence in
indica rice, Crop Sci. 28; 561-563.
Ponnuthurai, S., Virmani, S. S. and Vergara, B, S. 1984, Comparative studies on the growth
and grain yield of some rice {Oryza sativa L.) hybrids. Philipp. ]. Crop Sci. 9(3): 183-193.
Pradhan, S. B., Ratho, S. N. and Jachuck, P. J. 1990b. Development of new cytoplasmic genic
male sterile lines through indica x japónica hybridization. Euphytica 48; 215-218,
Pradhan, S. B., Ratho, S. N. and Jachuck, P. J. 1990a. Development of new cytoplasmic
genetic male sterile lines through indica x japónica hybridization in rice Euphytica
51: 127-130.
Rutger, J. N. and Shinjyo, C. 1980. Male sterility in rice and its potential use in breeding. In:
Innovative Approaches to Rice Breeding, IRRI, Manila, Philippines, pp. 53-66.
Sampath, S. and Mohanty, H.K. 1954. Cytology of semisterile rice hybrid. Cwrr. Sci, 2 3 :18JJ-
183.
Sasahara, T., Cui, C.H. and Kambayashi, M. 1986. Cold resistance in rice with some reference
to cytoplasmic effects. SABRAO 18(1): 69-71.
Satyanarayana, P.V., Kumar, I, and Reddy, M. S. S, 1995. A new source of thennosensitive
genic male sterility for two-line hybrid rice breeding. Int. Rice Res, Newslett. 20(1): 10.
Second, G. 1982. Origin of the genetic diversity of cultivated rice ( Oryza spp.)-. Study of the
polymorphism scored at 40 isozyme loci. ]pn. /, Gent. 57:25-57.
Senadhira, D. and Virmani, S. S. 1987. Survival of some F^ rice hybrids and their parents in
saline soil. Int. Rice Res. Newslett. 12(1): 14-15.
Shahi, B. B., Morishima, H. and Oka, H. 1.1969. A survey of variation in peroxidase, add
phosphatase, and esterase isozyme of wild and cultivated Oryza species, fpn. f. Genet. 44:
.808-319.

J.S. Nanda and S.S. Virmani 51
Shinjyo, C. 1975. Genetical studies oí cytoplasmic male sterility and fertility restoration in
rice, Oryza sativa L. Bull Coll Agrie., Lfnítí. Ryukus. 22:1-57.
Shirijyo, C. and Omura, T. 1966, Cytoplasmic male sterility in cultivated rice,. Oryza sativa L.
I. Fertilities of Fj, F2 and offsprings obtained from their mutual reciprocal backcrosses
and segregation of completely male sterile plants. Jpn. /. Breed. 16(Suppl. 1): 179-180.
Shinjyo, C. and Matomura, K, 1981. Inheritance of male sterility in isogenic lines of Taichung
65 possessing male sterile cytoplasm and fertility restorer genes from Oryza perertnts
W1080 strain, Jpn. J. Breed. 31 (Suppl. 1); 240-241, (in Japanese).
Shinjyo, C. and Sato, S. 1994. Chromosomal location of fertility-restoring gene Rf-2. Rice
Genet. Newslett. 11:93-95.
Shinjyo, C,, Ishimura, Y. and Tanaki, M. 1981. Inheritance of male sterility in isogenic lines of
Taicung 65 possessing male-sterile cytoplasm and fertility restoring genes of Oryza
perennis W1080 strain. }pn. f. Breed. 31 (Suppl. 1): 238-239 (in Japanese).
Shull, G. H. 1908. The composition of a field maize. Amer. Breed Assoc. Rep. 4:296-301.
Siddiq, É.A., Virmani, S.S., Ahmed, I.M. and Jachuck, P.J. 1996, Hybrid Rice-A promising
technological advancement for food security during 21st century. Paper presented at
the 3rd Inti Symp, Hybrid Rice, November 14-16, Hyderabad, India.
Singh, U.S. 1983. Hybrids excel standard cultivar. Deep Water Rice. 1:1.
Sprague, G. F. and Ebehart, S. A, 1977. Corn breeding. In: Corn and Corn Improvement, G.F,
Sprague, (ed)., Amer. Soc. Agron. Madison, WI, pp. 309-362.
Srivastava, M.N. and Seshu, D,V. 1983. Combining ability for yield and associated characters
in rice. Crop Scl 23:741-744.
Suh, H.S. 1985. Studies on multiple pistillate male sterile rice. Korean /. Breed. 17:439--443.
Suh, H.S, 1988. Multiple pistillate male sterile rices for hybrid seed production. In ; Hybrid
Rice, IRRI, Manila, Philippines, pp, 181-187.
Sun, Z. X., Min, S. K. and Xiong, Z. M. 1989. A temperature- sensitive mal^sterile line found
in rice. Rice Genet. Newslett. 6 :116-117.
Teng, L. S, and Shen, Z. T. 1994. Inheritance of fertility restoration for cytoplasmic male
sterility in rice. Rice Genet. Newslett, 11:95-97.
Tian, C., Cheng, X. and Liang, Z. 1980. Several views on popularization of Hsian (indica)
hybrid rice, Kunming Yunnan Nongye Keiji 2:12-18 (In Chinese).
Tran, D.T. and Nguyen, V.N. 1998. Global hybrid rice: Issues and challenges. Paper presented
19th Sess. Inti. Rice Comm., FAO, Cairo, Egypt, September 7-9.
Vaidyanath, K. and Reddy, G.M. 1985. Studies on genetic divergence in the genus Oryza.
J. Cytol Genet.20:59-68,
Vijaya Kumar, R. and Virmani, S.S. 1992, Wide compatibility in rice {Oryza sativa L.).
Euphytica 64: 71-80.
Virmani, S.S. 1996. Hybrid Rice. Adv. Agron. 57:577-^2.
Virmani, S.S. and Shinjyo, C. 1988. Current status of analysis and symbols for male sterile
cytoplasms and fertility restoring genes. Rice Genet. Newslett. 5:9-15.
Virmani, S. S. and Wan, B. H. 1988. Development of CMS lines in hybrid rice breeding, In:
Hybrid Rice, IRRI, Manila, Philippines, pp. 103-114.
Virmani, S. S. and Voc, P. C. 1991. Induction of photo- and thermo-sensitive male sterility in
indica rice. Agron. Ahstr. p.ll9.
Virmani, S.S., Khush, G.S. and Pingali, P.L, 1994, Hybrid rice for tropics; Potentials, research
priorities and policy issues. In: Hybrid Research and Development of Major Cereals in Asia
Pacific Region R.S.Paroda and M. Rai (eds.)., FAO, Bangkok, pp. 61-86.
Virmani, S. S., Young, J. B., Moon, H. P., Kumar, I. and Flinn, J. C. 1991. Increasing rice yields

Ricé Breeding and Genetics; Research Priorities and Challenges
Wan, J. M., Yanagihara, S., Kato, H. and Ikehashi, H. 1993. Multiple alleles at a new locus
causing hybrid sterility between Korean indica variety and a javanica variety in rice
(Oryzfl sativa L.) Jpn. J. Breed. 43:507-516.
Wang, C. L., Zou, J. S., Wang, Z. M., Li, C.G. and Li, H, B. 1991a. Identification of wide
compatibility and heterosis in rice, strain 02428 Xuan. Chinese /. Rice Sei. 51(1): 19-24.
(Chinese with English summary).
Wang, W.M. and Wen, H. C. 1995. New cytoplasmic male sterile line with lower negative
effects of cytoplasm on some quantitative traits in rice. Inti. Rice Res. Newslett, 20: 20 in
press.
Wang, X. M., Wang, M. Q., Mei, G. Z., Wu, H, Y., Duan, W. J. and Wang, W. J. 1991b.
Photoperiod-conditioned male sterility and its inheritance in rice. In: Rice Genetics II,
IRRI, Manila, Philippines, pp. 217-226.
Watnabe, Y. Sakaguchi, S. and Kudo, M. 1968. On the male sterile rice plant possessing the
cytoplasm of Burmese variety, Lead Rice, Jpn. J. Breed. 18 (suppl, 2): 77-78 (in Japanese),
Xu, X.J., Yin, H.Q. and Yin, H. 1991, Preliminary study of the temperature effect among S-1
and W6154S, Crop Res. (China) 5(2); 4-6.
Xiao, Y. H. 1981. Study on esterase isozyme of hybrid rice and three lines. Hubei Agrie, Sd, II.
(Chinese).
Xizhi, L, and Mao, C.X. 1994, Hybrid Rice in China-A Success Story. APAARI Publ.: 1994/3,
FAO Regional Office for Asia and the Pacific, Bangkok.
Yabuno, T. 1977. Genetic studies on the interspecific cytoplasm substitution lines of japónica
varieties of Oryza sativa L and O. glaberrima Steud. Euphytica 26; 451-463.
Yanagiharaa, S., Kato, H, and Ikehashi, H. 1992, A new locus for multiple alleles causing
hybrid sterility between an aus variety and javanica varieties in rice (Oryza satim L.).)pn.
¡.Breed. 42: 793-801.
Yang, R.C., Li, W.M., Wang, N.Y., lang, K.J. and Chen, Q. H. 1989. Discovery and preliminary
study on indica photosensitive genetic male germplasm 5460ps. Chinese J. Rice Set. 3(1):
47-48.
Yi, O. H., Shi, S.Y. and Jiang, J.R. 1984. Analysis of the esterase isozymes in three lines and F]
in Oryza sativa and prediction of heterosis. Acta Bot Sin. (Eng. Trans.) 26:506-512.
Yoshida, S.1981. Fundamentals of Rice Crop Science. Los Banos, Philippines.
Yoshimura, A., Iwata, N. and Omura, T. 1982. Linkage analysis by reciprocal translocation in
rice plants. III. Marker genes located on chromosomes 2,3,4, and 7 Jpn. I Breed. 32(4): 323-
332.
Yu, S.B., Li, J.X., Xu, C,G.,Tan, Y.F., Gao, Y.J., Li, X.H., Zhang, Qifa andSaghai Maroof, M.A.
1997. Importance of epislasis as the genetic basis of heterosis in an elite rice hybrid. 94:
9226-9231.
Yuan, L. P. 1977. The execution and theory of developing hybrid rice, Zhonggue Hongye Kexue
1; 27-31, (in Chinese).
Yuan, L.P., Li, Y.C, and Deng, H,D. 1990. Progress of studies in rice twin seedlings. Paper
presented 4th Annual Meeting Rockefeller Foundation's Inti. Program on Rice
Biotechnology, May 9-12,1990. IRRI, Manila, Philippines.
Yuan, L.P. and Fu, X.Q. 1996. Technology of Hybrid Rice Production. FAO, Rome. 84 pp. ,
Zhou, T.B., Xiao, H.C., Lei, D.Y. and Duan, Q.X. 1988. The breeding of indica photosensitive
male sterile line. /. Hunan Agrie. Sei. 6:16-18.

Sustainable Integrated
Rice Production
S. V. Shastry^ D. V. Tran^, V, N. Nguyen® and J, S. Nanda*
INTRODUCTION
Rice provides about two-thirds of the caloric intake for more than two
billion people in Asia, and one-third of the caloric intake of nearly one
billion people in Africa and Latin America. Rice production in the world
has trebled since the turn of the century, which has enabled a stable
decline in the world rice price (Mitchell, 1987). The growth in rice
production was particularly rapid during the green revolution after the
introduction of high-yielding varieties (HYVs) such as Taichung Native
1, ADT27, H4, H5, and IRS (Chandler, 1979).
World population, however, continues to grow at a high rate of
about 1.7% annually (most of which occurs in the developing world),
resulting in almost 90 million more consumers of agricultural products
per year. In 1992, world demand for milled rice was about 390 Mt and is
expected to increase about 2% per annum over the next two decades
(Yap, 1992). However, the phenomenal growth in rice production
achieved with the adoption of green revolution technologies in areas with
high yield potential, notably in Asia, has shown signs of accumulative
^Former-Director of Research, IITA, Ibadan, Nigeria.
Senior Rice Agronomist, FAO.
Rice Agronomist, FAO, Rome.
Rice Breeding and Genetics Specialist, GUY/91/001, FAO of the United Nations, Rome

The term "green revolution" is employed to describe the phenomenal
growth in cereal production (including rice) that was achieved in the
1960s in areas with a high potential for yield, by concurrent application of
genetic, fertilizer, irrigation and plant protection technologies. In the case
of rice, the green revolution has enabled a steady increase in production
in Asia, which is well ahead of population growth in many countries
(Singh in FAO, 1992). In Latin America, it allowed the region to be nearly
self-sufficient, producing about 97 % of its consumption (FAO, 1993). In
Africa, however, it has not been able to cope with an ever-increasing high
rate of consumption.
The green revolution has undoubtedly established the technical
feasibility of maintaining high production in irrigated ecology, but has
made a limited contribution in areas with poor water control and totally
failed in areas with problem soils. In areas where it has been successful,
the green revolution has introduced the type of rice farming that demands
a high investment. However, significant concerns have centered on
"mining" the soil for plant nutrients, changes in the status of rice pests
from minor to major economic importance, the negative impact on the
environment, and on the long-term sustainability of production growth.
The widespread use of HYVs has reduced the biodiversity and the germ
sfnrk of fradifinnal nrndiirHon svstems. Recent!v. vield decline has been
54 Rice Breeding and Genetics; Research Priorities and Challenges
stresses (De Datta, 1994) and has generated numerous problems with
aspects ranging from the biological to environmental and socioeconomic
(Tran and Ton That, 1994). At the other end of the scale, the increase in
rice production by expansion of upland area, notably in Africa, has
accelerated deforestation and desertification which, in many countries,
has caused severe losses in natural resources and biodiversity, as well as
irreparable damages to the environment.
International concern about rice production, therefore, has shifted
from per se to sustainability of production growth, from the profitability
of rice farming as an enterprise to conservation of the resource base and
maintenance of the integrity of the environment for prosperity. Every
technological opportunity for expanding rice production and elevating
rice productivity is being scrutinized for sustainability over a longer time
frame and for its impact on the environment. In the coming decades, the
challenge facing rice production will not be in expanding rice production
and productivity at any cost, but in integrating the diverse elements of
technologies and evaluating the sustainability of technology packages.
THE GREEN REVOLUTION

SUSTAINABLE IRRIGATED RICE
S.V. Shastry et at.
not to lose the tempo gained by the green revolution, while at the same
time mitigating the associated negative effects. There is a constant need to
keep the demand and supply situation of rice in view while investigating
the tolerance limits for existing technology packages and emphasizing
the need for innovation in specific areas.
SUSTAINABLE RICE PRODUCTION
Most rice production systems, with the possible exception of slash-andburn
shifting rice cultivation, are considered relatively sustainable. These
systems continue, albeit at a low level of output per unit of land and time.
However, Rutten (1988) stated that if the concept of sustainability serves
as a guide, it must include the use of technology and practices that
enhance and at the same time sustain productivity. The output of a
sustainable system, which is different from the output of a stable system,
must not only remain unchanged, but must increase with time to satisfy
the demand of an increasing population. A sustainable agricultural
system is one that can evolve indefinitely towards greater human utility,
greater efficiency of resource use, and a balance with the environment
that is favorable to humans and to most other species (Harwood, 1988).
The broad definition of sustainable development was elaborated by FAO
as follows (FAO, 1989):
Sustainable rural development is the management and conservation
of the natural resource base, and the orientation of technological and
institutional change in such a manner as to assure the attainment and
continued satisfaction of human needs for the present and future
generations. Such sustainable development, in the agriculture, forestry
and fisheries sectors, conserves land, water, plant and animal genetic
resources, is environmentally non-degrading, technically appropriate,
economically viable and socially acceptable.
Sustainable rice production systems, therefore, may be defined as
the ones in which technology packages are chosen for increasing paddy
yield with a rational concern for resource, economy, and the
environment. Such systems are exercisable, although they do not yet
exist in ready forms.
Rice is grown over a wide range of climatic, soil, and water regimes,
and different classifications exist; the one adopted by the International
Rice Research Institute (IRRI, 1984) is used here,

■' ■
Wtä
..--■■■
56 Rice Breeding and Genetics: Research Priorities and Challenges
(IRRI^ 1993). The irrigated rice ecology has experienced the most
intensive adoption of green revolution technologies, such as HYVs,
inorganic fertilizers, and pesticide application. In addition, the level of
crop intensification is highest in irrigated rice ecology. The
intensification of irrigated rice production, however, has led to a longterm
decline in productivity. Although evidence is still scarce, yields of
some modern varieties under optimal management at IRRI and other
locations in the Philippines have declined by 0.1 to 0.3 the t ha”^y'^ over
a 20-year period (Flinn, et al., 1982).
The intensive monoculture of rice in an irrigated ecology has led to a
build-up of salinity and waterlogging, micronutrient deficiencies,
formation of hardpan, and an increased pest build-up. Other major
concerns relating to sustainability of irrigated rice production systems
are the increasingly limited water supply available for rice production,
the clogging of waterways with aquatic weeds, and the plateau in yield
potential of conventionally bred rice varieties of ttie green revolution
generations (IRS type).
The sustainability of rice production in irrigated ecology, however,
will be increasingly dependent upon yield increase and cropping
intensity, as there is limited scope for the expansion of net area. During
the past decade, rice, paddy areas in most of the major rice producing
cormtries in Asia had remained static or even declined. In Latin America
and Africa, although there are regions with areas still potentially suitable
for conversion to irrigated rice production, the cost of development of
an irrigation infrastructure is too high to be affordable for many
countries. The average yields of irrigated rice in many countries are still
only about 3 to 4 t ha'^. There are therefore ample opportunities for
increasing yield. It is necessary to mobilUe the inputs required to adopt
recommended practices fully and thus exploit judiciously the
production potential of irrigated lands.
WATER MANAGEMENT
The availability of water for irrigated rice productiori will become scarce
increasingly as water in reservoirs, which were built for irrigation
purposes, is increasingly diverted for household and industrial uses. The
situation is particularly critical in arid and semi-arid areas. The efficiency
of water management practices in rice production must be critically
L----£ :—
r m i i a f l^ p v i p W f i d

Classic methods of maintaining soil productivity are fallowing, crop
rotation, and application of organic manures. With the advent of
fertilizers for farming, the classic methods of soil productivity
maintenance receded into the background. Major plant nutrients, which
used to limit crop production, such as nitrogen, phosphorus and
potassium, could now be enriched by application of fertilizers, which
were high-analysis compounds. Genetic improvement of crop plants,
which paid particular attention to crop cultivars responding to high
rates of fertilizer application, further elevated the role of fertilizers in
farming. However, several developments in the 1970s contributed to
dampening the enthusiasm for fertilizers. After the energy crisis of 1973,
the prices of fertilizers soared, resulting in an unfavorable balance
between the costs of grain and fertilizers. Concerns about the
environment have also conflicted with the promotion of fertilizers.
Leachates from fertilized soil end up as nitrates in waterways.
The concept of fertilizer application needs to be guided by the
__ --t *1 J 1-------i.----------------------------X
S.V. Shastry et at. 5 7
an excessive waste of water, especially when feeder canals are unlined
and field bunds are neither well constructed nor maintained. Recent
research has shown that intermittent flooding to keep the soil saturated
provides better water-use efficiency (Kandiah, et al., 1990). Integrating
the irrigation schedule with the time of fertilizer application and
weeding can avoid unnecessary drainage, thereby saving water. The
timing of irrigation following rainfall distribution also reduces water
loss through field runoff.
Much water is wasted, since the water tariff is based on the area
irrigated rather than on the volume of water used. Volumetric fees would
improve the farmers' incentive to aim for better water-use efficiency.
Pivotal to all reform in water use is to educate beneficiaries on the need to
reform and on the resultant opportunities, such as:
• adherence to a tight calendar of field operations so as to minimize
wasteful water runoff;
• consolidation of holdings and compensatory land use;
• choice of early-maturing cultivars and improvement of cropping
intensity;
• organization of community nurseries;
• maintenance of field channels and drains;
• timely and efficient weed control.
INTEGRATED NUTRIENT MANAGEMENT

58 Rice Breeding and Genetics: Research Priorities and Challenges
long term. The notion of balanced fertilization envisages total nutrients
available in the soil summed with and added fertilizers. In practice,
however, the balance is often influenced by the nutrient composition of
added fertilizer,s. The recommended rates of P and K may either be
inadequate or unnecessary. During 1981-1983, balanced fertilization
produced higher yields and economic returns in demonstrations
conducted in different agro-ecological zones in Bangladesh (Roy and
Pederson, 1990).
Every means to improve the efficiency of fertilizer use has become
important. Technologies of the prefertilizer era, such as green manuring
and the application of farmyard manure and compost, are being
reemphasized. Advantage is being taken of the possibilities for biological
nitrogen fixation with blue-green algae, Azolla ferns and free-living
bacteria. Attempts to minimize fertilizer losses from the soil by modifying
the practice of fertilizer application are underway. Practices such as
coating urea with organic wastes and mixing it with soil to form pellets
have gained renewed appeal.
Water management needs to be modified so as to compleiuent the
use of fertilizers. Knowledge of crop plants, such as their growth rates
peak demands for nutrients and, is essential if the best benefits from
fertilizers are to be realized while agronomic practices that synergize
with management of soil, water, symbionts and crop plants will also
have to be employed. The integrated nutrient management system
(INMS) is therefore a composite package of improved crop agronomy to
increase and sustain crop production without "mining'^ the soil and / or
polluting the environment.
INTEGRATED PEST MANAGEMENT
"Pest" is a collective term employed for various living organisms (e.g.
weeds, insects, fungi, viruses and rodents) that cause crop yield losses
as a result of competition, infection, infestation or ingestion. Experience
with chemical and genetic control and other approaches has led to a
consensus that pests are not universally destructive and that their
annihilation is not always desirable. A pesticide is an alleviator of specific
stresses. Furthermore, it is often uneconomical to reduce pest population
to very low levels. For an intelligent management of pests, it is necessary
to recognize the conditions which promote crop growth and performance
as well as the ecology and population dynamics of pests and their natural
enemies.

S;V. Shastry et at. 59
expense of weed growth. Choosing cultivars with a rapid tillering habit,
coupled with practices such as planting of young seedlings with close
spacing to accelerate tillering, can produce a dense crop canopy that can
smother weeds. This effect may be further enhanced when the cultivar
has drooping leaves and when the crop has been adequately fertilized.
An integrated package of genetic, mechanical, water and fertilizer
technologies can thus produce a sea change in the crop-weed balance.
Management of major rice diseases caused by fungi (rice blast),
bacteria (leaf blight) and viruses (tungro, grassy stunt) is achieved by a
combination of practices. The choice of a highly resistant genotype is
beginning to lose its appeal because such resistance is often contained in a
limited number of major genes and is therefore frequently ephemeral.
Instead, durable resistance controlled by polygenes is preferable. Altering
the sowing times so as to minimize the chances of synchrony of the most
susceptible host stage with environments-favorable for disease build-up,
control of vectors or cultivation of trap crops for vectors of viral disease,
and chemical control guided by disease forecasting (e.g. for blast) have
visible effects on slowing down the epiphytotics. With the change in the
"plant type" cultivars, the status of some diseases has also changed. For
example, bacterial leaf blight was recognized as a major disease only after
the semidwarf varieties came to be grown in monsoon-affected Asia.
A blanket cover of insecticide protection (i.e., prophylaxis) has been
progressively recognized as being more harmful than useful, since it
eliminates a wide array of beneificial insects. Whereas genetic resistance
has offered dramatic opportunities with respect to gall midge and brown
planthopper, biotic variation in these insects has necessitated a constant
vigil. The importance of crop growth and health in partially
compensating for insect infestation is convincing in the context of stem
borer and indicative in the context of other insects. For example, the
damage caused by borers during the tillering phase (reflected in dead
hearts) is almost fully compensated and so, in most cases, chemical
control of borers at this stage is unwarranted.
A new movement, integrated pest management (IPM), has gained
currency in many developing countries. The emphasis of IPM is on
experimentation at the field level and it is aimed at farmers. IPM views
the crop in a holistic way and seeks to give weight at once to crop
agronomy and plant protection. Emphasis is given to field diagnosis of
the problem in its incipient stage. The complexity of mixed infections
and/or infestations is discussed and farmers are educated about the role
of helpful fauna in rice fields. They are also taught to "dodge" the use of
pesticide and to decide for themselves when it is absolntplv nerpRR^rv to

60 Rice Breeding and Genetics: Research Priorities and Challenges
MECHANIZATION
Irrigated rice farming is in itself labor intensive and labor input rises
further with the level of intensification; so mechanization of a sort that
matches the resource endowments of farmers is essential. Rice farming
mechanization has made a wide range of labor-saving equipment
available for irrigated rice production. However^ developing countries
have little option other than to pay attention to intermediate scale
mechanization and place greater reliance on renewable sources of energy.
The humid tropical and semiarid belt of sub-Saharan Africa is one such
example. The gap between the hand hoe and machete level of
mechanization and tractorization is so wide that the former is grossly
inadequate and the latter is woefully unsustainable. The promotion of
draft animals for traction and transport need not be considered
negatively, but rather recognized as the most appropriate in its particular
context. Mechanization in rice production must be viewed not only as a
means to reduce labor inputs and human drudgery, but also a way to
generate employment through the local manufacture and maintenance of
equipment.
GENETIC IMPROVEMENT
Wl
The key to green revolution technologies is HYVs. However, the yielding
potential of conventionally bred varieties of the green revolution
generations for irrigated rice has reached a plateau. The development
and adoption of new generation HYVs , especially with short duration,
are still needed for sustainable irrigated rice production. High-yielding
and short-duration rice varieties permit not only a higher level of rice
production, but also better crop diversification, as less land is needed
for a shorter time to produce a certain quantity of rice. Diversification of
intensified rice monoculture -reduces the risk of increasing the water
table, which causes salinity in arid areas, and waterlogging in humid
areas. The development of hybrid rice with a yield potential of about 15
to 30% over the best yields of conventional varieties gives much hope for
sustainable irrigated rice production (Ton That, 1992). In China, the
wide adoption of hybrid rice for cultivation has saved more than 2
million ha of land for diversified crops, fisheries and livestock
production. India, the Philippines, Vietnam and other countries are also
engaging in the development and production of hybrid rice. Efforts to
develop super HYVs is another move in the right direction.
Farmers have bred their own rice varieties over the centuries. Today,
the wide adoption of HYVs is the main cause of genetic erosion, as

S.V. Shastry et at. 61
traditional rice is replaced and natural biodiversity declines. National
governments and international organizations should seek appropriate
approaches to protect genetic diversity in time. They should enhance the
conservation, evaluation, exchange and utilization of germplasm at the
local, regional, national, and international level, while taking farmers'
rights into consideration. The wide-spread use of HYVs has also reduced
fish production horn the traditional rice-fish culture system in many
Asian countries, as the agronomic: conditions used for HYVs are not
suitable for fish life (Choudhury, 1996). The IPM approach and the
greening of rice production by encouraging organic fertilizer use would
promote the resurgence of fish, frogs, and many phytons in rice
microfauna and microflora, resulting both in improved farmers' incomes
and human nutrition.
DEVELOPMENT AND MANAGEMENT OF IRRIGATION
SCHEMES
Most discussion regarding irrigation centers on the scale of projects. The
so-called large irrigation schemes are losing their popularity for a variety
of reasons. Water-use efficiency is discouragingly low for a given level of
investment, since a major portion of water loss is traceable to storage and
transmission, The loss in transmission is high because, in an effort to
scale down cost, many countries sacrifice sophistication when designing
transmission canals, land-shaping, and making provision for drainage.
The rehabilitation of large-scale irrigation schemes is an expensive
operation.
Therefore, the so-called small-scale irrigation schemes have grown
in importance. The smallest irrigation scheme is the harnessing of
underground or surface water for lift irrigation of an area totally owned
by a family. Next in the scale are the medium-sized>'reservoirs which rely
on the principle of water harvesting found in dry zones in Sri Lanka and
India or in inland swamp valleys in West Africa. Larger schemes entail
the diversification of rivulets for gravity flow. The long-term
sustainability of these schemes depends on the recharge of aquifers and
the precipitation in a watershed; it is impaired by lowering of the water
table, silting of reservoirs and clogging of waterways. The management
of these irrigation schemes needs to be vigilant and correctives must be
introduced on a timely basis.
There are some pervasive problems in rice farming in small
irrigation schemes. The small size of holdings, along with their dispersal
over wide distances, leads to unavoidable losses in water. When the
sowing of seedbeds is unduly protracted, the efficiency of water use

62 Rice Breeding and Genetics; Research Priorities and Challenges
suffers because of an "idle run" in irrigation canals. The tendency of
farmers to raise a separate nursery on their own is a waste of resources.
A consensus to consolidate the nurseries (i.e., community nurseries)
close to the reservoir would contribute to the efficiency of water use.
The desilting of reservoirs and the transportation of soil to rice fields
need to be encouraged. Holdings in easy reach of the canal tend to be
favored against the tail enders, who suffer from delayed planting as well
as submersion after transplanting. A reversal of insequence plantings
(early planting in low areas) would improve the aggregate performance
of a scheme.
APPROPRIATE RICE FARMING SYSTEMS
The diversity in rice farming systems is largely due to variation in the
physical/ biological and human factors that influence farming. In
Bangladesh, for example, the rice crop is grown in three seasons-rspring,
autumn and winter-and farmers allocate resources for farming activities
by looking at each rice-growing season as a farm unit, guided by such
considerations as subsistence, insurance against crop failure, income
generation, and avoidance of peak period for labor. Over the last three
decades, the Boro (spring) season has grown in importance. The
uncertainty of production from the Aman (winter) rice due to flooding
has made farmers realize the value of spring rice, for which HYV
technology is admirably suitable.
The rice-based cropping system has emerged as an important
program in South and South Asia. A vast resource of rice fallow is being
mobilized to raise supplementary income-generating crops such as grain
legumes, sunflower, and peanut. The monocrop farming scenario of
monsoon-affected Asia is rapidly changing. The underlying incentives
for these changes are increased pressure on land, declining margins with
monocrop farming, and an increased awareness of the sustainability of
rice multicrop systems.
Crop intensification is the farmer's practical measure in densely
populated areas, such as the Red River delta and the central plain in
Vietnam. The traditional crop production system is double cropping of
irrigated rice, using high levels of both internal and external inputs.
Large irrigated lands are left idle between two crops during the cool
weather starting from late September to February, Sweet potatoes,
potatoes and winter vegetables are traditionally grown on limited areas

S.V. Shastry et at. 63
The HYV technology has, in some cases, accelerated changes in the
farming system. Availability of an early-maturing HYV such as ADT27
has triggered a large-scale conversion of single-cropped (Samba) rice
lands into double-cropped (Kurivai and Thaladi) rice lands in the
Thanjavur delta of Tamil Nadu, India. In parts of northern Andhra
Pradesh, two crops of an early-maturing HYV are being grown in areas
where one crop of a late-maturing variety (GEB 24) used to be grown.
Several nontraditional rice-growing areas have emerged, (e.g. Pimjab,
Haryana and Telengana in India) in the process of replacing other crops—
maize, sorghum, millet and cotton—with HYVs of rice.
Since the 1920s, pasture in irrigated rice rotations has been
undertaken in New South Wales in Australia and still offers potential to
substantially improve the productivity, profitability, and sustainability
of a rice cropping system. This system has helped Australian rice farmers
to obtain high yields (about 1 0 to 1 2 t ha"^) with less use of chemical
fertilizers (only 60 to 100 kg of N ha'^) and has provided them with
opportunities for farm income through complementary animal industries.
Farmers usually grow two rice crops, which are then followed by three
consecutive crops of subterranean clover {Trifolium subterraneaum).
These legume-based pastures supply subsequent rice crops with fixed
soil nitrogen, improve soil structure, and break weed cycles. However,
research still needs to address the problems of disease, waterlogging,
salinity, and acidity as well as the need for improved pasture
establishment and management technology in order to conserve the
ecology and maximize production (Tran in FAO, 1994a).
Current interest in diversification is targeted towards utilizing
renewable sources of energy more efficiently, minimizing the use of
scarce high-energy inputs, conserving the productivity potential of
natural resources, alleviating socioeconomic constraints, and lessening
the negative impact of farming on the environment at large. The objective
is to ensure the sustainability of the production system over a long time
frame. The model aims to maximize the opportunities for recycling
products, ensure that the low-cost by-products of one subenterprise
become the inputs of another subenterprise, and avoid entropy and
waste.
The most obvious aspect of diversification is the complement
between the raising of crops and livestock, with Ihe former contributing
the feed and the latter providing the manure. The less obvious benefit
from diversified farming of crops and livestock is a more even
distribution of labor demand, a better dispersal of income generation,

64 Rice Breeding and Genetics: Research Priorities and Challenges
crops into cereals, legumes and vegetables, and livestock into large and
small ruminants, nonruminants and poultry) acts as an additional
insurance against total failure of the enterprise that may result from
unforeseen fluctuation in the demand for and price of farm products.
The by-products of rice farming, such as straw, bran and broken
rice, are often wasted unless the rice farmers keep livestock, dairy
animals, and poultry. The use of rice field bunds for raising vegetables
and the use of rice fallow for grain legumes would minimize farmers''
trekking back and forth between vegetable gardens and rice fields,
which is done in West Africa, with resultant entropy. The use of bunds
and fallow would also encourage better utilization of by-products in pig
rearing, as is done in China. There, interest in rice-cum-fish culture has
been revived, particularly in shallow-flooded rice farms where the use
of pesticides is not widespread. Traditional rice-fish culture systems are
commonly exploited in rice fields with water depth of less than 50 cm. It
is estimated that if only 5% of irrigated rice areas with a production
target of 300 kg ha'^y'^ and 15% of deepwater rice areas with 600 kg
hay'^ were used for rice-fish culture, a total of 3.2 Mt of fish could be
produced (Choudhury, 1996),
SUSTAINABLE RAINFED LOWLAND RICE
Rainfed lowland rice, including deepwater and tidal wetlands,
constitutes about 31% of the world's harvested rice areas and 21% of
world rice production (IRRI, 1993). The soil surface in the rainfed
lowlands is protected from horizontal displacement by a layer of flooded
water saved from rainfall. In Asia, rainfed lowland rice areas come close
to irrigated rice in terms of area. The yield potential of shallow rainfed
areas with assured rainfall is as good as that of irrigated areas, but it has
remained underexploited. These areas provide great opportunities for
increased rice production. Flooded soils can be continuously farmed.
Submersion cuts down the severity of weed competition and the soil
reduction that follows flooding helps mobilize the plant nutrients in the
soil into an available form, notably of iron and phosphorus. Free-living
bacteria living on the surface of rice roots thrive on the oxygen pumped
by the foliage of flooded rice; some of them are known to fix atmospheric
nitrogen. On the negative side, the decomposition of organic matter
under anaerobic conditions can result in products that are toxic to the
rice plant. Flooded rice itself is also a contributor to the build-up of
greenhouse gases, particularly methane.

S.V. Shastry et at. 65
iron and phosphorus. A depth of up to 30 cm can render fertilizers less
efficient. A depth in excess of 50 cm tends to activate the elongation of
stemS/ and only special cultivars can withstand a water depth of more
than 1 m. Some general principles apply to water management in all the
lowland rice ecosystems, namely, fields should never be allowed to dry to
the extent of developing cracks, since major losses of nitrogen can occur
and weed competition can be intense; fields should not be flooded to
more than 30 cm in depth during the tillering phase; and fields should be
drained if possible when fertilizer is applied (so as to enable absorption
into soil) and reflooded one or two days later to conserve the nitrogen
added to the reduced zone.
Lands which benefit from horizontal seepage from adjacent ridges
and for this reason remain wet or flooded for a considerable period after
rains stop are highly productive for rice. Their soils are less prone to
erosion and less vulnerable in drought than the corresponding uplands.
However, fluctuation in the water table and alternation between dry
and wet surfaces in hydromorphic soils favor aquatic and upland weeds,
while the loss of soil nitrogen is accelerated by nitrification and leaching.
Minor improvements, such as constructing field bunds to follow contour
lines and making provision for drainage, can elevate the productivity of
hydromorphic lands. Improvement of hydromorphic soils in inland
swamps for rice production is wide in Asia. In Africa, similar efforts have
been initiated in West African countries such as Burkina Faso, Benin and
Sierra Leone.
There are about 138 Mha of wetlands suitable for the cultivation of
bunded rainfed lowland rice in tropical Africa (Ton That, 1982), but only
about 1.5 % of these are actually cultivated with either rainfed lowland or
irrigated rice. Development of these wetlands for rice cultivation
increases the sustainability of rice production in Africa and helps to slow
down the clearing of forest for upland rice cultivation.
In Africa, Asia and Latin America, large areas of deepwater and
mangroves have been successfully exploited over the years for
agricultural production, including rice. Population pressure, however,
has put this exploitation in environmental danger. Several ecosystems
among those converted to agriculture have shown themselves to be
unsuitable for sustainable deepwater rice production and have suffered
. from degradation and abandonment. Such wastelands can be restored in
an integrated land development plan at a reasonable cost and effort.
Adequate land-use planning and surveys including socioeconomic and
environmental considerations for the short and long term should
orecede the restoration of these abandoned lands. Several

66 Rice Breeding and Genetics: Research Priorities and Challenges
SUSTAINABLE UPLAND RICE-BASED CROPPING
Upland rice (or dryland rice) grows on about 20.4 Mha, which represents
about 14 % of the'world^s rice area. Numerous subsistence farmers grow
upland rice, mostly on poor, well-drained soil with an erratic rainfall and
under shifting or permanent cultivation, or as a pioneer crop. Rice is
grown either as a monocrop or in crop mixtures in the montane regions
of the Far East, in the humid tropical belt of Africa and the subhumid
savarmas of Africa and Latin America, with average yields ranging from
1 to 1.5 t ha'^ (Coulter, 1994). These areas which are marginal for rice
production are expected to be grown to rice in the medium term, as long
as the production from lowlands fails to meet the demand. The major
problems of the sustainability of upland rice production in the humid
zone include soil erosion, soil acidity, deficiency of nutrients, and weed
infestation in intensive cultivation. The conventional solution to this
problem is to practice arable cropping for a short cycle in partially cleared
landscapes and to turn them over to bush (bush fallow) for a longer
period. With mounting population pressure on land, fallow periods are
progressively reduced and soil physical properties are rendered
unsuitable for arable farming. Improper soil management has led to soil
erosion, degradation and deforestation in many parts of the world.
The principal measure to improve upland rice cropping systems is
water and soil conservation. Water conservation should be emphasized
with techniques of minimum tillage, bunding, weeding, proper plant
deirsity, and mulching and contour cultivation. The principle for soil
erosion control is to provide a continuous vegetal cover throughout the
year to diminish rainfall intensity and runoff (Pande, Tran and Ton That
inFAO,1994).
The alley cropping system is a recent innovation whereby the
perennial bush vegetation is replaced by leguminous shrubs which are
planted in organized rows between which food crops are raised (Kang,
et al, 1990). The composition of the alleys could eventually be modified
so as to generate income. Alley cropping permits a continuous
cultivation of arable crops, with the alleys providing the mulching
materials where they are most needed. Mulch minimizes the direct
impact of rain on the fragile tropical soil. On decomposition, the organic
residue supplies nitrogen to the soil. Green leaf manuring and alley
cropping are just as relevant for the production of small ruminants
(sheep and goats) as for rice in the hydromorphic and freely drained rice
lands of tropical Africa. Many of the leguminous shrubs and trees are

The integrated approach to rice farming must satisfy both economic and
ecological considerations, which balance the costs at farm level with
those for society, The economic viability of technology will determine its
acceptance by farmers. When technology is also environment friendly, it
minimizes the costs to society. Improved technology permits the
substitution of the more expensive resources by the less expensive.
Therefore, application of improved technology tends to bring down the
unit cost of output in a production system and/or permit the expansion
and diversification of enterprises.
The efficiency of a rice production system is determined by the
productivity of one or more of the following major factors: land, water,
and labor. The relative importance of these factors varies from location to
location. For example, the productivity of land is an important
consideration in the Asian hiunid and subhumid tropics where land is
scarce. In semiarid Asia and Sahel countries, the productivity of water
and land assume an equal importance. In Asian countries where rural
employment is scarce, the productivity of labor assumes less importance
than in countries of West Africa where labor scarcity is acute. The
productivity of water is rarely taken into account when evaluating rice
farVmolno-ips Althniiah irrigation is reallv. an exoensive
S.y. Shastry et at. 67
Crop rotation is an age-old method of maintaining soil fertility and
is extremely relevant for upland rice. The improved rice-pasture system
in Latin America (i.e., one year of rice and three years of pasture)
involves block or strip rotation: an upland field is divided into several
blocks or strips, each of which is cultivated with an upland crop,
including upland rice, in a given year. In West Africa, the establishment
of a food crop legume such as Vigna spp. in association with or after
upland rice is a promising system for sustainable upland rice cultivation.
Progress in rice research in this agroecology has been slowly achieved
at the national and international level. Fimdamental physiological
features of upland rice should be studied to shed more light on the work
being done in varietal improvement and agronomy. More productive and
profitable upland rice farming systems must adapt to local environments
as substitutes for slash-and-bum cultivation. Subsistence living leaves
small farmers little room for taking risks, New approaches are needed,
including national policy reorientation and politicial will, in order to
stabilize and reduce vulnerable upland rice areas and make them more
economic, productive and sustainable when exploited (Tran, 1986).
SOCIOECONOMIC VIABILITY

68 Rice Breeding and Genetics: Research Priorities and Challenges
risk-prone areas. It is only in exceptional circumstances (e.g. in
groundwater exploitation) that farmers bear the real cost of irrigation.
As a consequence, water-use efficiency in rice farming is appallingly
low. The sustainability of many irrigation projects has, therefore, been
seriously questioned.
A major technological breakthrough is the adaptation of semidwarf,
plant-type cultivars to the tropics. The plant architectural attributes of
the semidwarf gene in the variety Deo-Geo-Woo-Gen have contributed
to an improved interception of solar radiation by the tropical rice plant.
This is perhaps the most significant low-cost and environment-friendly
technological advance in rice farming this century.
TECHNOLOGICAL ORIENTATION AND RISK MANAGEMENT
Advances in basic sciences have opened up fresh avenues for the
development of new technologies. Research organization has assumed an
interdisciplinary task-force approach to resolve a problem and innovate
solutions. Initially, emphasis was on the twinning of physical and
biological sciences. Towards the end of this century, this interdisciplinary
ring was extended to economists, anthropologists and land environmentalists.
Application of production technologies culminated in the
green revolution, which brought into focus the awareness and concerns
for issues such as risks in production, equity in income distribution,
economic sustainability, and impact on the environment.
Conceptualization of design, generation, transfer and evaluation of
technologies has thus emerged as a collective responsibility of biology,
physics, and social scientists. There is a need for further refinement of
the farming systems approach to research through incorporation of
macroeconomic goals, scrutiny of economic sustainability, and sensitivity
to environmental issues. This reorientation arises from the fact that an
improvement in technology often has system-wide implications and
affects public policy and investment. In addition, rice development
cannot be considered in isolation but needs to be placed in the perspective
of overall economic development.
Farming as an enterprise has always had to reckon with and reconcile
an element of risk. The magnitude of what is at risk increases with every
incremental change in production. The HYV technology can be shown to
be robust in an agronomic sense due to its stable performance over a wide
range of climatic and managerial situations. Yet it does not automatically
follow that the risk involved in adopting HYV technology is minimal. In
reality, the risk taken by farmers using HYVs is proportional to the yield

The high growth rate in rice production during the green revolution was
made possible by the appropriate measures taken by international and
national authorities. Sustainable rice production, can also be achieved,
therefore, with appropriate policies to protect natural resources and
mobilize human resources and capital for a rational exploitation of
existing lands and new land reclamation. Policy measures such as
taxation of resource utilization and price regulation create a favorable
market environment for sustainable rice production. For example, the
imposition of land-use taxes coupled with incentive prices for forestry
and fishery products encourage the shift from rice production in
unfavorable areas such as upland and mangrove lands to agroforestry,
wildlife reserves, and fishery activities. The application of volumetric
fees for irrigation water will encourage farmers to maintain irrigation
canals and economize in water use. Favorable market prices and
conditions encourage farmers to adopt appropriate production
techniques and farming systems.
The need for sustainable agricultural production is more important
in China than anywhere else. Chinese scientists, development officials,
politicians and farmers have responded to this need with concerted
'"phasing" approaches; in each phase various technologies have been
integrated into packages. The first phase was initiated in 1950, at which
time the following measures were popularized: use of improved varieties;
growing of strong and healthy seedlings; intensive land preparation;
application of proper plant densities; balanced fertilizer application;
rational irrigation; and pest and disease control. During the second
Hpvelonment of double cropping rice and the concepts of three
S.V. Shastry et at. 69
magnitude of yield losses (in actual, not percentage terms) that
epidemics and epiphytotics can inflict on a higher-level production
system is much higher than that of traditional low-production systems.
The foregoing discussion does not imply an abandonment of existing
research establishment de novo of institutions mandated to conduct
research on environment-friendly and economically sustainable technologies.
A change in orientation is what is needed. As previously stated,
there are no new technological leads available which meet the new
goals. What is initially expected is that there will be a shift to the
conventional technologies and practices which went out of fashion with
the advent of seemingly inexpensive chemical and mechanical technologies.
This may be a transient phase, which then gives way to major
breakthroughs from biotechnological research.
NATIONAL POLICIES FOR SUSTAINABLE RICE PRODUCTION

At the practical level, sustainable rice production is possible with
consistent and concrete programs and projects which take into account
both the welfare of farmers and the conservation of natural resources.
Rice production projects such as swamp rice development, reclamation of
mangrove rice areas and improvement of rice systems (e.g. deepwater,
rainfed lowland, upland and irrigated) should be planned and
implemented with the following priorities:
• Consideration of the sociopolitico-cultural setting, economic
environment and ecology to ensure the feasibility, profitability,
and sustainability of the program and its acceptance by farmers.
• Conformity with current national policies and laws (if any) aimed
at harmonizing economic returns and resource conservation. If
such policies and laws do not exist, they should be established.
• Promotion of the application of technologies requiring less fossil
energy and imported inputs and, at the same time, encouragement
of the utilization of local resources and renewable energy which
aim to help farmers become self-reliant in production activities and
70 Rice Breeding and Genetics: Research Priorities and Challenges
crops a year and high-yielding crops were emphasized. By noting the
low outputs of individual technologies^ emphasis was placed on an
integrated approach encompassing land development, fertilization,
improved cultivation and improved seed and cropping systems during
the third phase. During the fourth phase, in 1980, hybrid rice and hybrid
maize were popularized, with emphasis on multiple cropping and use
of improved seed-growing techniques. The fifth phase, in 1985,
prompted establishment of market correction measures. These
integrated policy measures have increased crop yield significantly and
enabled China to be not only self-sufficient, but also an exporter of rice
(Wang in FAO, 1990).
Similarly, in Indonesia integrated policies comprising the
introduction of proven technologies, together with agricultural
extension, input distribution, banking services and cooperatives, have
been the key factors in successful sustainable rice production (Dudung,
1994).
In the long term, sustainable rice production needs comprehensive
taxation law and price regulation to prevent undesirable influences from
other sectors of the economy; for example, the silting of reservoirs by
deforestation of watershed areas, the dumping of toxic materials from
industrial and mining sectors in agricultural lands and water, and the
unchecked expansion of housing construction on prime irrigated lands.
FORMULATION OF PROGRAMS AND PROJECTS

S.V. Shastry et at. 71
• Establishment of indicators and frameworks for information
collection^ relating to the monitoring and evaluation of project
implementation as well as the integrity of the environment and the
population's welfare as rice development and population
progress.
CONCLUSIONS
The application of science and technology to rice farming has so far
progressed in the direction of relying on nonrenewable natural resources
such as fossil fuel, of which there is not an unlimited supply. Therefore,
the pattern of development that has given prosperity to many developed
and some developing countries is unlikely to be accessible to those
countries which have yet to modernize and improve the efficiency of rice
farming. There is a need in all rice-growing countries to improve the
efficiency of agrochemicals and to modify production packages so as to
give greater attention to renewable sources of energy. Many of the crop
production and protection technologies that prevailed in the prefertilizer,
prepesticide era need to be reexamined in this changed context.
Rice production should not be viewed in isolation but as a part of
holistic farming systems in which the farmer's income and welfare as
well as the diversity of the social, biological, and physical environment
should be integrated into the design of appropriate technologies. The
intensified monocropping rice systems are still in need of diversification
to maximize the opportunities to recycle products, avoid entropy and
waste and rejuvenate degraded rice soils caused by continuous flooding.
Technological advancements are useful when accompanied by
appropriate national policies, which are supported by consistent and
concrete programs. Therefore, in the long run sustainable rice production
requires the formulation and implementation of relevant programs for
rice research, development, and production.
FAO has a crucial role to play in articulating the paradigm shift in
rice farming, clarifying the sustainability goals and ecological concerns,
and presenting a range of technological, institutional and managerial
options that enable its member cotmtries to meet the current needs for
rice without compromising the requirements of future generations.
References
Barghouti, L.G. and Umali, D. In: World Bank Technical Paper No. 180, pp. 107-126.
Washington, DC, World Bank.
Chandler, R.F. 1979. Rice m the Tropics; A Guide to the Development of a National Program.

72 Rice Breeding and Génetics: Research Priorities and Challenges
Choudhury, P,C. 1996. Integrated rice-fish culture in Asia with special reference to
deepwater rice. In: Proc. 18th Session IRC, 5-9 September 1994, Rome.
Coulter, J.K. 1994. Upland rice production and its environmental impacts. IRC Newsl., 39; 41-
49.
De Datta, S.K. 1994. Sustainable rice production: challenges and opportunities. IRC Newsl.,
39: 209-219.
Dialla, B.E. 1994. The adoption of soil conservation practices in Burkina Faso. Indigenous
Know. Dev, Mon. 1:10.
Dudung, A.A. 1994. Modernizing rice farming; Indonesia's experience in conducting the
green revolution, IRC Newsl 39; 265-275.
FAO, 1989. Sustainable agricultural production: implications for international agricultural
research, FAO Research and Technology Paper No. 4, Rome.
FAO. 1990, Technology for sustainable crop production and agricultural development in
China. A consultancy report submitted to FAO/RAPA, 47 pp.
FAO. 1992, Research and development strategies for the increase and sustained production
of rice in Asia and the Pacific region, FAO/RAPA; 1992/17.
FAO. 1993, FAO Production Yearbook. 1992. Rome.
FAO. 1994a, Mission report on the conference Temperate Rice: Achievement and Potential,
Yanco, NSW, Australia, 21-24 February 1994, Rome.,
FAO. 1994b. Improved Upland Rice Farming Systems. Rome, 117 pp.
Flirm, J.C., De Datta, S.K. and Labadan, E. 1982, An analysis of long-term rice yield in a wetland
soil. Field Crop Res., 5: 201^216.
Harwood, R.R. 1988. The contribution of agro-ecology to development of a sustainable
agriculture. Paper presented at the colloquy Completing the Food Chain: Multisectoral
Strategies for Combating Hunger Malnutrition. Smithsonian Institute, Washington,
DC, 4 October 1988.
IRRI. 1984. Terminology for the rice growing environment. Manila.
IRRI. 1993. Rice in crucial environments (IRRI 1992-1993). Manila, 65 pp.
Kandiah, A,, Ton That, T. and Carpenter , A.J. 1990. Area changes and system choices for
irrigation in Asia 1990-2000. IRC Newsl., 39: 23-25.
Kang, B.T., Reynolds, L. and Atta-Krah, A.N. 1990. Alley cropping .Adv. Agron. 43:315-359.
Lacy, J. 1994. Rice check: a collaborative learning for increasing productivity. Abstract
presented at the conference Temperate Rice: Achievement and Potential, Yanco, NSW,
Australia, 21-24 February 1994.
Mitchell, D .0 .1987. Rice market prospects'to the year 2000. Paper presented at the seminar
Recent and Future Movement in World Rice Prices, Jakarta, Indonesia, 13-14 January
1987,
Roy, R.N, and Pederson, 0,S, 1990. Economic use of fertilizer for rice. IRC Newsl, 39:118-
126,
Ruttan, V.W. 1988. Sustainability is not enough. Notes presented at the symposium Creating
Sustainable Agriculture for the Future, 30 April 1988, University of Minnesota.
Ton That, T. 1982. Potential and constraints of rainfed lowland rice development in tropical
Africa. IRC Newsl, 31(1): 1-6.
Ton That, T. 1992. Le programme de riz hybrid a la FAO, IRC Newsl, 41:51-59.
Tran, D.V. 1986. An overview of upland rice in the world. In: Progress in Upland Rice Research,
Manila IRRI p, 51-66.
Tran, D.V. and Ton That, T. 1994. Second generation of high-yielding rice varieties. IRC
Newsl,39:127-132.
Yao, C.L, 1992. Situation of the world rice market in 1991. IRC Newsl, 41:65-70.

* International Rice Research Institute, P.O. Box 933,1099, Manila, Philippines.
4
Drought and Submergence
in Rice Production
Osamu Ito"^, Gloria Cabuslay"^ and Evangelina Ella*
INTRODUCTION
Rice is grown in various agroecosystems defined on the basis of
hydrology. The rice ecosystem can be roughly classified into four types:
irrigated, rainfed lowland, upland, and deepwater. More than half of
the rice lands are irrigated and nearly three-quarters of the rice
production comes from this ecosystem. The irrigated ecosystem
provides a micro-environment which is most favourable to rice plants
and allows stable yields over time. If irrigation is not available, rice
plants would fully depend on water supply from rainfall and would
often suffer from water deficit due to erratic rainfall, which makes grain
production unstable and unpredictable. On the other hand, if the
drainage system is not well developed, rice plants are partially or fully
submerged by water for variable periods of time depending on the
intensity of rainfall. The deficit and excess of water (drought and
submergence) are two major abiotic constraints which limit rice
production significantly when rice cultivation moves out from the
irrigated ecosystem.
Drought, defined as a period of no rainfall or no irrigation that
affects crop growth (Hanson et al, 1995), has long been recognized as the
primary constraint to rainfed rice production (De Data et al., 1975}

Rice Breeding and Genetics: Research Priorities and Challenges
Mackill et at, 1996). Drought causes a reduction in biomass and
consequently^ a reduction in yield. Yield losses are more severe when
drought occurs during the reproductive phase (Sarkarimg et ah, 1995).
Flooding is a serious constraint to plant growth and survival in rainfed
lowland and deepwater areas because excessive water results in partial
or complete submergence of the plant. Partial submergence occurs when
40-99% of the shoot is under water (Setter et ah, 1987), Adverse effects of
flooding cover approximately 2,2 Mha of rice land which includes 1.5
Mha of flash-flood areas of rainfed lowland rice and 5 Mha of deepwater
rice (Khush^ 1984). Depending on the water depth and duration^
flooding in rice fields may be categorized into two types (Ram et ah,
1996): (i) flash or intermittent and (ii) stagnant or prolonged.
I. CHARACTERISTICS OF DROUGHT ENVIRONMENTS
There is a wide range of water-stress environments in rainfed ricegrowing
areas, differing in both timing and intensity of water stress
(Fukai and Cooper, 1995). These environments can be grouped into two
major categories, namely upland and rainfed lowland. The rainfed
lowland ecosysteny accounts for 27% of the total rice area in the world
(Sarkarung et ah, 1995). In South and Southeast Asia, rice is grown on
approximately 40 Mha with yields averaging only 1.5 t ha'^ (Wade et ah,
1995). On the other hand, upland rice grows on about 17 Mha worldwide
with upland areas contributing only 4% to total world rice production
(IRRI, 1996).
A. Upland
Upland areas may have deep soils with high extractable soil water
content, although water-stress development is generally more severe
there than in lowland areas (Fukai and Cooper, 1995). Furthermore,
upland areas are characterized by extreme diversity of soils and
topography. The complexity of the ecosystem is well represented by
cultivation of widely differing traditional cultivars with a variety of
farming practices (IRRI, 1990).
The major problems involving upland rice cultivation can be
grouped into three categories (IRRI, 1990):
(1 ) environmental problems, e.g., soil degradation, deforestation,
lack of capital to purchase inputs, and poor access to markets;
(2 ) crop problems, e.g., poor management, lack of efficient tools, high
weed infestation; and
(3 ) plant problems, e.g., low yield potential, drought, and blast
damage.

Osamu Ito, Gloria Cabuslay and Evangelina Ella 75
B. Rainfed Lowland
Rainfed lowland areas may be further grouped based on the presence or
absence of a hardpan layer in the soil. The hardpan which develops as a
consequence of puddling and construction of bunds in lowland
conditions results in better retention of surface water, which delays
development of plant water stress. In areas where puddling is not
practised or where the soil is sandy, a hardpan may not develop,
resulting in a large percolation loss of rain water (Fukai and Cooper,
1995).
As a result of hydrologic conditions which may fluctuate from
submergence to drought, various systems of crop establishment are
employed in rainfed lowland, from direct dryseeding to transplanting
(Wade et ah, 1995). Harvesting a rice crop is largely determined by the
time at which the monsoon arrives and the amount of rain that falls
(Mackill et ah, 1996).
Drought is manifested by the uncertain onset of rains at rice sowing
or transplanting and by prolonged dry periods during the reproductive
phase. Farmers practice direct seeding in their upper rice-fields using
traditional, drought-tolerant, medium-statured cultivars to get an early
harvest even if a mid-season drought occurs.
Aside from hydrologic conditions, other problems which affect crop
growth in the area are the occurrence of diseases such as blast in rice
(Fukai and Cooper, 1995) and low soil fertility (Mackill et ah, 1996), since
farmers minimize inputs because the risk of water-related stress is high.
II. EFFECT OF DROUGHT STRESS ON PLANT GROWTH
A. Effect on Morphological Characters
1. SHOOT
A general effect of drought stress is a reduction in size of plants (stunting
or growth retardation). The height of the plant is particularly affected
(Laude, 1971). According to (Kramer, 1969), leaf area, cell size, and
intercellular volume usually decrease, while cutinization, hairiness, vein
density, stomatal frequency, and thickness of both palisade layer and
entire leaves usually increase. The amount of epicuticular wax, a
significant component of the cuticle, is higher in dryland-adapated rices
than in irrigated rices. This often results in relatively thick, leathery,
highly cutinized foliage, generally described as xeromorphic.
In rice, leaf rolling and death of leaves are criteria found useful in
assessing levels of drought tolerance in a large-scale screening (Chang

Rice Breeding and Genetics; Research Priorities and Challenges
et al, 1974). Singh et al. (1995) reported that drought imposition stopped
leaf elongation when the soil moisture tensions exceeded 40-60 kPa in
all establishment methods studied. Leaf rolling was found to reduce leaf
area index by almost 50% to compared unrolled leaves. This was the
primary cause for decline in transpiration rate in the initial stages of
drought imposition. Further decline in tran^iration was caused by
stomatal closure at soil water tensions of 10^ kPa irrespective of the
establishment method.
Studies on root-related characteristics have elicited much interest from
scientists because it is through roots that rice plants take up moisture,
and a wide range of varietal differences exists in the rice root system
(Mackill et al., 1996). Deeper roots may permit access to water that is not
available to other plants with a shallower root system (Moreshet et ah,
1996). Furthermore, drought stress causes pronounced changes in root
structure, such as increased branching (Yamauchi et ah, 1996) and
density (Eghball and Maranville, 1993). However, most root studies
focus only on morphology and ignore dynamic physiological activities
in roots, e.g. respiration. Root respiration merits attention because it is
closely linked to metabolic processes and uptake of water and nutrients
(Ito et ah, 1996). The role of root respiration in adaptation of plants to
drought stress has not yet been fully explored.
Drought-affected plants generally exhibit a small root system
configuration and in many cases the reduction in size of root system is
directly proportional to the magnitude of water shortage (Yamauchi et
ah, 1996). Lilley and Fukai (1994a) found that root growth ceased in all
the rice cultivars studied when water deficit was imposed at either the
vegetative or reproductive stage. Cruz et ah (1986) reported that water
stress decreased root length of IR54, and this was attributed to increased
soil mechanical impedance. Morita and Abe (1996) observed that roots
of upland rice respond to drought conditions with an increase in degree
of branching.
According to Slayter (1973), two types of effects of water deficit on
root development can be expected, the first being a reduction in rates of
meristematic activity and root elongation directly associated with the
level of internal water deficit; the second, an effect of suberization on the
water and nutrient uptake properties of the root system as a whole. As
rates of root elongation are reduced, the rate of suberization exceeds the
rate of elongation, and the nonsuberized zone is reduced, until it is
virtually eliminated in nonelongating roots. This phenomenon, common
under conditions of severe water stress, substantially reduces the
effective surface of the roots and their activity as absorbing organs.

Osamu ItO/ Gloria Cabuslay and Evangelina Ella 77
B. Effect on Yield
Lilley and Fukai (1994b) found that water deficit during vegetative
growth caused an insignificant reduction in grain yield but tended to
delay panicle initiation. On the other hand, water deficit imposed during
the reproductive period reduced grain yield to 20-70% of the irrigated
control. A slow growth rate during panicle development reduced grain
number and potential grain size, while cultivars which recovered
quickly after water deficit had a relatively higher grain yield.
Singh and Ingram (1991) reported that water stress during booting
to early grain filling caused the greatest yield losses of 77% in IR20, IR46,
and IR72. Stress treatments during three different growth stages
reduced plant height, culms per plant, leaf area, grain yield components,
grain yield and daily as well as seasonal évapotranspiration.
In a study using 20 early-maturing rice cultivars, severe drought
stress prolonged the maturity period of all cultivars by 2-27 days
(Dikshit et ah, 1987). The long dry period and the prolonged maturity
period reduced the grain yield by 10-91%. Significant correlation was
found between maturity prolongation and yield reduction due to
drought stress.
According to Dey and Upadhyaya (1996), of the three critical stages
of growth—seedling, vegetative and anthesis—drought during anthesis
is the most serious and devastating to yields because of its adverse
effects on pollination and the flowers become sterile. The decreased
sugar delivery to reproductive tissues upon inhibition of photosynthesis
due to drought stress triggers metabolic lesions leading to failure of
male gametophyte development (Saini, 1997).
C. Effect on Plant Functions
1. P h o t o s y n t h e s is
Photosynthesis is the driving force of plant productivity. The ability to
maintain the rate of photosynthetic CO2 fixation under environmental
stresses is fundamental to the maintenance of plant growth and
production (Lawlor, 1995). Reduced biomass results in a small
photosynthetic area, leading to a reduced assimilate storage in
vegetative organs which ultimately limits potential grain yield even if
favourable conditions return (Begg, 1980).
Drought has short-term as well as long-term aftereffects on
photosynthesis (Boyer and Mcpherson, 1976). In the short-term,
photosynthesis may be affected by changes at the chloroplast level and/
or by stomatal movement. Two kinds of aftereffects appear following

Rice Breeding and Genetics; Research Priorities and Challenges
rewatering of the plants. First, there may be incomplete recovery of leafwater
potential, which causes photosynthesis to remain below the levels
of the control. This appears to be caused by breaks in water columns or
other modifications of the pathway for water transport in the plant.
Second, there may be a direct aftereffect of drought on the
photosynthetic process. Both depend on the severity of desiccation: the
more severe desication is, the more severe are its aftereffects.
Jones (1981) reported that soil moisture stress causing stomatal
closure for periods of up to 8 days during the critical period from 2 0
days before flowering to 1 0 days after flowering has little effect on the
percentage of filled grain. Longer periods of stomatal closure can
seriously affect yields.
At the biochemical level, reports point to the effect of drought stress
on photosystems, more on PSII than PSI. He et al. (1995) found that
damage to PSII photochemistry in waterstressed wheat leaves is brought
on by a decreased rate of synthesis and increased degradation of PSII
proteins.
In a review, Lawlor (1995) suggested the following sequence of
events when C3 leaves are exposed to increasing water deficit:
(a) Mild stress or partial loss (25%) of turgor and a small decrease in
relative water content or RWC (100% down to 90%) has little effect on
photosynthetic metabolism. At this stage, the main limitation in
photosynthesis is the reduction in diffusive conductance caused by
stomatal closure and decreased Q (intercellular CO2 concentration).
Accumulation of sucrose and starch may occur, reflecting the
maintenance of a positive balance between synthesis and consumption
despite reduced photosynthesis.
(b) Further loss of turgor and reduction of RWC below 85%
decreases stomatal conductance and potential CO2 assimilation rate due
to metabolic alterations, but Cj may continue to decrease. The low Cj
leads to an increase of the In vivo oxygenase/carboxylase ratio of
Rubisco, causing a larger relative flux of carbon through the
photorespiratory glycolate pathway. Increased photorespiration,
relative to photosynthesis, recycles CO2 , consuming relatively more
NADPH than ATP.
(c) Light-harvesting, electron transport and reduction of pyridine
nucleotides and other electron acceptors are little affected in the
physiological range of stress. The reduced to oxidized pyridine
nucleotide ratio and the size of the reduced pyridine nucleotide pool
increase with moderate to servere stress.
(d) ATP content decreases with moderate to severe stress due to
impaired synthesis of ATP by the coupling factor, which is inhibited by
the ionic conditions in the chloroplast. Consequently the ATP/reduced
pyridine nucleotide ratio falls.

Osamu Ito, Gloria Cabuslay and Evangelina Ella 79
(e) Synthesis of RuBP is inhibited as a result of limited ATP supply^
so the potential for CO2 assimilation and photorespiration decreases
progressively with developing stress and control shifts from CO2
availability to RuBP synthesis.
(f) Abnormal regulation of, or damage to enzymes is not the major
lesion in photosynthetic metabolism.
(g) CO2 assimilation and synthesis of trióse phosphate, sucrose and
starch decreases substantially while consumption of sucrose by
respiration continues so that the total carbohydrate content of the leaves
falls, starch more so than sucrose.
(h) The proportion of electron flow to O2 (Mehler reaction) increases,
generating superoxide and hydrogen peroxide which damage
membranes and enzymes. A greater proportion of the energy is
dissipated by qj.jp (non-photochemical quenching) than by qp (quenching
due to photochemistry). Dissipation mechanisms (carotenoids,
including the xanthophyll cycle and antioxidant systems) become
increasingly important, removing energy and reductant and destroying
toxic compounds generated as energy is transferred to unphysiological
acceptors.
(i) Severe water deficit causes substantial inhibition of
photophosphorylation and a further decrease in ATP content. CO2
assimilation and photophosphorylation almost stop, respiratory
processes dominate, and C¡ rises greatly.
(j) Excessive energy loads on the thylakoids and the detoxification
systems eventually lead to membrane damage and irreversible loss of
photosynthesis. These processes may be linked to the accumulation of
metabolites such as proline.
2. C a r b o n P a r t it io n in g
The term partitioning projects the concept of a central resource pool that
is distributed among sinks (Dingkuhn and Kropff, 1996). Partitioning of
assimilates changes in the course of phenological development as
different organs are formed. In modern high-yielding varieties, the dry
weight of the root system is similar to the shoot at the seedling stage but
is only 1 0 % that of the shoot at flowering, and even less at maturity.
Boyer and McPherson (1976) found that maize can utilize previously
accumulated dry matter for translocation to the grain, and suggested
that photosynthetic activity before as well as during the grain-filling
period was the important determinant of grain yield during drought.
The translocation mechanism, while having less photosynthate available
for transport, was itself relatively unaffected.
The two stable endproducts of photosynthesis are sucrose and
starch. Sucrose is synthesized in the cytoplasm and starch in the

Rice Breeding and Genetics: Research Priorities and Challenges
chloroplast of mesophyll cells. According to Van der Werf (1996)/
primary regulation of leaf carbon partitioning between sucrose and
starch is believed to reside in the cytosol. Considering the low carbon
supply under drought conditions, a shift in chemical partitioning of
carbon occurs in favour of sucrose accumulation or starch
remobilization in the leaf cells of stressed plants. Continued sucrose
synthesis is therefore required for export while sucrose accumulation is
essential for osmotic adjustment during stress.
Van der Werf (1996) further stated that low carbon supplies coupled
with the need for some carbon to be used to osmotic adjustment rather
than growth/storage would necessitate modifications in sink demand
and alterations in shoot-to-root ratio and source-to-sink relations. Most
often changes in the ratio are in favor of root growth and reduced sink
load, both of which are reflected in increased root-to-shoot ratio. Ma
and Lu (1990) observed promotion of root growth in hybrid rice by mild
water stress, probably coupled with improved oxygen supply.
In wheat, Nicolas et al. (1985) reported that under control conditions,
the distribution of carbon among ear, stem, and roots did not change
during the experimental period. The grains acted as the main sink with
daily carbon increment 3 times that of the stem and 4 to 6 times that of
the roots. Under mild drought stress, the sink strength of the roots
increased relative to that of the grains and stem in the more tolerant
cultivar. However, under more severe stress, roots did not compete well
with grains.
3 . W a t e r -u s e e f f ic ie n c y (W U E )
Evapotranspiration studies on rice revealed that 759-1150 kg of water,
with a mean of 875 kg, were required to produce 1 kg of rough rice grain
(Shih et al, 1982). Considering that this resource is limiting in rainfed
environments, a desirable trait for a rice cultivar to maintain growth and
yield in drought-prone areas is to have efficient water use. Water-use
efficiency of the whole plant is usually defined as the total dry matter
produced per unit of water used (Boyer, 1996). At the leaf level WUE
(also termed transpiration efficiency) is defined as the ratio of the rates
of net carbon assimilation to transpiration (Van den Boogard et al,
1995). WUE of tropical upland rice was improved by stomatal closure
and leaf rolling during mild and intermediate water stress (Dingkuhn
et al, 1989).
Transpiration efficiency is indirectly measured as carbon isotope
discrimination (A), based on the relative contents of and
plant tissues. The hypothesis is that cells accumulate relatively more
than because ^^C0 2 is lighter, hence diffuses more rapidly than

Osamu ItO/ Gloria Cabuslay and Evangelina Ella 81
^^C0 2 and is fixed more rapidly by Rubisco (Boyer, 1996). The unused C
diffuses out according to the extent of stomatal opening. The inward
diffusion and use of ^ ^ 0 2 correlates with photosynthesis and dry mass
while the outward diffusion of ^^C0 2 correlates with transpiration. Thus
the relative uptake of and correlates with water-use efficiency.
This development renewed interest in water-use efficiency studies
which usually involved rigorous daily weight measurements. With the
availability of mass spectrometers to measure relative contents of ^^C0 2
and ^^C0 2 , WUE can now be determined with ease even under field
conditions, as plant parts can be harvested and immediately analyzed
for relative uptakes of ^^C0 2 and ^^C0 2 -
Farquhar and Richards (1984) reported that plants with high WUE
had high ratios of to or less discrimination against
According to Boyer (1996), substantial water limitation usually gives a
negative relationship between discrimination and WUE but under
relatively favorable conditions, the relationship tends to become less
negative or even positive. Among 28 upland rices with different wateruse
efficiencies, Dingkuhn et aL (1991) reported carbon isotope
discrimination to correlate negatively with WUE across all cultivars and
within japónica and aus groüps, but not among indica rices. Comparison
of experimental data with varietal screening results of the International
Rice Research Institute (IRRI) revealed that cultivars with good seedling
vigor had high WUE and low discrimination.
4. R o o t r e s p ir a t io n
Roots appear to be the poor relations when it comes to the allocation of a
limited supply of photosynthate (Wardlaw, 1990). A decrease in root
growth is commonly mentioned as the primary result of drought stress
(Pardales and Kono, 1990). This might lead to a reduced respiratory
activity in the roots since rate of root respiration was found to correlate
positively with the relative growth rate of roots (Poorter et al., 1991).
Indeed, exposure to a dry soil led to a gradual decline in root respiration
of Triticum aestivum, predominantly due to the engagement of the
alternative pathway (Nicolas et at, 1985). The decline in respiration
correlates with the accumulation of organic solutes,
Plant mitochondria possess a branched electron transport chain that
contains two pathways: the cytochrome pathway and the alternative
pathway (Atkin et a/., 1995). Both the cyanide-sensitive, SHAM
(salicylhydroxamic acid)-resistant cytochrome pathway and the
cyanide-resistant, SHAM-sensitive alternative pathway obtain their
electrons from Qj. (reduced ubiquinone). However, in contrast to the
cytochrome pathway, electron transport from Q, to O2 via the alternative
pathway does not lead to the synthesis of ATP. In addition, Millar et al

Rice Breeding and Genetics: Research Priorities and Challenges
(1995) stated that the alternative pathway releases energy as heat which
cannot be utilized for any metabolic activity. Induction of the alternative
pathway is generally found under stress conditions^, e.g. wounding of
tissue, chilling, osmotic stress, and drought (Wagner and Krab, 1995). In
wheat, although total root respiration decreased imder drought, less
respiration took place via the alternative pathway, especially for the
more tolerant cultivars (Nicolas et al, 1985).
From titrations with specific inhibitors (especially cyanide and
hydroxamates), it was concluded that the cytochrome pathway is used
preferentially and that the alternative pathway acts as an overflow, only
engaged when the cytochrome pathway is saturated (Wagner and Krab,
1995). Thus engagement of the alternative pathway in roots tends to
decrease when the availability for respiration decreases and increases
when the demand for carbohydrates for other processes decreases
■(Lambers ei al.,. 1991). However, recent studies suggest that the
cytochrome pathway does not have to be saturated for the alternative
pathway to be engaged; hence these studies are incompatible with the
Bahr and Bonner (1973) hypothesis of a preferential pathway (Millar et
al., 1995; Dry ei al., 1989).
III. MECHANISMS OF DROUGHT RESISTANCE
Drought resistance refers to those properties which enable plants of a
given genotype to grow and reproduce (yield) normally under drought
conditions (Chang et al, 1974), To achieve this, plants develop certain
morphological and physiological traits which adapt them better to water
stress in rainfed areas. This is supported by the existence of "traditional"
upland and rainfed lowland cultivars in the rice germplasm. Scientists
have been trying for years to exploit these adaptive mechanisms in an
attempt to increase yields in rainfed environments. These mechanisms
can be classified into escape, ayoidance, and tolerance mechanisms.
A. Escape Mechanism
According to OToole and Chang (1978), drought escape is the most
effective adaptive mechanism as far as productivity is concerned. It
aims at crop growth during the period of high rainfall and high soil
water availability to escape the drought period (Fukai and Cooper,
1995). One way of achieving this is to grow short-duration cultivars
which tend to escape late-season drought (Wonprasaid et al., 1996).
Another method is to use photoperiod-sensitive cultivars whose
sensitive reproductive stages are photoperiodically controlled to

Osamu Ito, Gloria Cabuslay and Evangelina Ella 83
coincide with the peak monsoon season, allowing the crop to complete
grain filling under Ihe adequate water regime (O'Toole and Chang,
1978). Most traditional rainfed lowland rices are sensitive to
photoperiod (Mackill et al., 1996).
B. Avoidance Mechanism
Strategies for avoiding low plant water status during a drought period
include extracting more water from the soil and using soil water slowly
during the early stages of a drought period so that more is available in
the later period (Fukai and Cooper, 1995). Some root and shoot
characteristics that enhance uptake and conservation of water during a
rainless period are leaf rolling, stomatal closure, thick cuticle
development, prolific root system, and deep roots (O'Toole and Chang,
1978; Samson et al, 1995).
C. Tolerance
Tolerance of water stress, usually involves the development of low
osmotic potentials, which characterize many plant species found in arid
environments (Morgan, 1984). Osmotic adjustment, or accumulation of
solutes by cells, is a process by which water potential can be decreased
without an accompanying decrease in turgor (Taiz and Zeiger, 1991).
According to Taiz and Zeiger (1991), most of the osmotic adjustment
can usually be accounted for by increases in concentration of a variety of
common solutes, including sugars, organic acids, and ions (especially
K^). However, a high concentration of ions can be severely inhibitory to
enzymes so that this occurs mainly within the vacuoles, where the ions
are kept out of contact with enzymes in the cytosol or subcellular
organelles. Because of this compartmentation of ions, some other solutes
must accumulate in the cytoplasm to maintain water potential
equilibrium within the cell. These other solutes, called compatible
solutes, are organic compounds that do not interfere with enzyme
functions. Proline is a commonly accumulated compatible solute; others
are sugar alcohol, sorbitol, and a quaternary amine, glycine betaine.
Nicolas et al (1985) found osmotic adjustment to be higher in roots of
drought-tolerant wheat cultivar, with contribution four times that of
sugars.
Despite much research on the accumulation of proline and other
compatible solutes during stress, it remains unclear whether these
accumulations are resistance mechanisms or symptoms of stress
(Mackill, 1996). In general, scientists believe that these are consequences

84 Rice Breeding and Genetics: Research Priorities and Challenges
of drought injury and, therefore, selection for these has not been proven
useful
IV. ENVIRONMENTAL CHARACTERIZATION OF
FLOODWATER
The degree of flooding is determined by many variables such as rainfall
pattern and intensity, topography of the location, soil properties, the
drainage system and so on. Not only does the depth of floodwater but
also its physicochemical characteristics (O2 and CO2 concentration, pH,
turbidity etc.) greatly affect plant survival during submergence. It has
been noticed that in order to obtain consistent genotypic differences in
the germplasm improvement trials for submergence tolerance, much
more attention should be paid to the floodwater characteristics. Field
research data on submergence are not interpretable and an experiment
may not be repeatable because of insufficient biophysical information
about the environmental conditions of floodwater. Levitt (1980) cited
several mechanisms important in flooding of soils (referred to as
waterlogging), but did not include the additional effects associated with
partial and complete submergence. There are several factors involved
with the adverse effects of flooding in rice. Some of these factors may
combine, e.g., siltation of leaves may affect other factors such as
mechanical damage, light, and gas diffusion. Limited gas diffusion is the
most important factor during flooding (Setter et ah, 1995). Gas diffusion
is 10^-fold slower in solution than in air (Armstrong, 1979). Any gas
produced under the water during flooding (e.g. O2 during the day, CO2
at night, and ethylene) increases its concentration at the production site
whereas any gas consumed (e.g. O2 at night and CO2 during the day)
decreases its concentration at the consumption site due to the slow gas
diffusion in water. This is supported by several data on gas
concentrations obtained in rice field floodwater (Setter et ah, 1988a and
Setter e.t ah, 1988b). A limited CO2 supply is common in flood-prone
environments.
The CO2 concentration in floodwater during turbulent flash floods
tends to be in equilibrium with that in air due to rapid mixing (Setter et
ah, 1987) but once the water level stabilizes, floodwater may become
stagnant, resulting in low CO2 concentration due to large iDoundary
layer effects (Smith and Walker, 1980; as discussed by Setter et ah, 1989).
Some areas in Thailand have floodwater with relatively high pH, which
rises above 7.0 during the day presumably due to CO2 consumption
through photosynthesis (Setter et ah, 1987). Coupled with little
turbulence in the floodwater and the resistance of the boundary layer
facing the gas envelope around the leaves would lead to low CO2 , high
O2 , and high pH in the boundary layer (Setter et ah, 1989).

Osamu ItO/ Gloria Cabuslay and Evangelina Ella 85
In some environments^ O 2 is completely absent (anoxia) in the
fioodwater particularly during the night when the O2 produced in the
daytime has been consumed for respiration. Burg and Thimann (1959)
reported that an ethylene precursor accumulated during anoxia and that
in waterlogged tomato plants it was transported from anoxic roots to
the shoots via root-shoot vasculature (Jackson and Campbelb 1975a and
1976). When the precursor reached the aerated shoots^ a large Amount of
ethylene Qackson and Campbell 1975a, 1975b and 1976) and faster rate
of ethylene production were observed (Bradford and Dilley^ 1978;
Jackson, 1979; Jackson et al,, 1978) after 24-72 h waterlogging, especially
at warmer temperatures (Burg and Thimann, 1959; Jackson, 1985).
However, in submerged photosynthetic tissues, the O2 deficiency may
not occur due to O2 evolution via photosynthesis during the day when
the fioodwater becomes supersaturated with O2 (Heckman 1979; Setter
et at, 1987 for Thailand; Whitton et al., 1988 for Bangladesh; Sinhababu et
a l, 1991; Setter et at, 1995 for India). In principle, this high O 2
concentration during the day may be injurious due to elevated
photorespiration or oxidative damage associated with postanoxic injury,
i.e., reentry of air following exposure to anoxia.
Another important limitation to photosynthesis under water is some
flash-flood affected areas is low irracliance due to turbidity. There was a
reduction to < 1% irradiance in air at only 40 cm depth in one floodaffected
location in eastern India (Setter et at, 1995). Whitton et al. (1998)
made the same observation in fioodwater in Bangladesh due to surface
algal floes.
V. EFFECT OF SUBMERGENCE STRESS ON PLANT GROWTH
A. Morphology
1. A e r e n c h y m a f o r m a t io n
An internal, longitudinally connected gas-filled intercellular space
formed by cell separation or by breakdown of cortical cells or pericycle
(aerenchyma) promotes root growth and survival in oxygen-deficient
conditions (Ap Rees and Wilson, 1984; Armstrong, 1979).
Such gas-filled channels would allow rapid gas movement from
shoot to root apex to supply O2 for root respiration and to release O2 into
the rhizosphere for oxygenation. Oxygenation is of importance because
it helps detoxify chemically reduced iron, manganese, and hydrogen
sulfide (Gambrell et al, 1991) and may support nitrifying bacteria in the
conversion of ammonia to nitrate (Blom et al, 1994). Aerenchyma not
only promotes long-distance oxygen transport to the roots, but also

86 Rice Breeding and Genetics: Research Priorities and Challenges
facilitates the connterflow of volatile compounds that accumulate in the
anaerobic soil and plant tissues which include ethanol (Crawford and
Finegan, 1989)^ CO2 , and methane (Vartapetian and Jacksort; 1997).
High porosity in primary and adventitious roots is achieved by the
formation of aerenchyma^ although cubic as opposed to hexagonal
packing also increases porosity (Yamasaki, 1952; Justin and Armstrong,
1987). Aerenchyma is formed either by selective cell lysis commonly
found in deeper rooted wetland species (Justin and Armstrong, 1987) or
by cell separation and differential rates of expansion, and both
developed constitutively in most but not all the wetland and amphibious
species examined (Smirnoff and Crawford, 1983), The proportion of
root occupied by aerenchyma in some wetland and amphibious species
such as Rumex crispus (Laan et ah, 1989), rice (Justin and Armstrong,
1987), and willow Qackson and Attwood, 1996) is further promoted by
poor aeration. The formation of aerenchyma is the most important
adaptive feature in facilitating O2 movement down into the roots. The
movement of O2 through the aerenchyma is usually by diffusion though
other mechanisms may occur as described by Armstrong et ah (1996). A
comprehensive study of the types of aerenchyma in wetland plants
(including rice) and nonwetland plants has been done by Justin and
Armstrong (1987).
The size of aerenchyma in a root depends on several factors and
their interactions. Aerenchyma is present in many roots of rice
genotypes regardless of the environmental factors (Luxmoore and
Stolzy, 1969) but its volume varies with the genotypes (Barber et ah,
1962; Justin and Armstrong, 1991). Aerenchyma formation in roots of
other cereals, however, is very slow in well-aerated conditions but
greatly enhanced during waterlogging, e.g., in stagnant and partial
shortage of oxygen or in anoxic solutions (for corn and wheat, Yu et ah,
1969; Varade et ah, 1970; Drew et ah, 1979). This response is more
evident in lines with higher waterlogging tolerance (Huang et ah, 1994;
Huang and Johnson, 1995). In corn it was shown that such response is
mediated by ethylene (Drew et ah, 1979) which forms in significantly
increasing amounts when the oxygen level is about 2 0 % of air saturation
(Brailsford et ah, 1993). Low O2 stimulates ethylene synthesis in the roots
(Jackson, 1985; Atwell et ah, 1988). Stagnant conditions lead to the
entrapment of ethylene produced in the plant tissues (Jackson, 1985).
Ethylene action induces programed cell death in cortical cells of primary
roots in association with a disorientation of microtubules in cells
destined to collapse (Baluska et ah, 1993), degradation of the cell wall
(Webb and Jackson, 1986), and an increase in cellulase activity (Jackson
et ah, 1993; He et ah, 1994).

Osamu Ito, Gloria Cabuslay and Evangelina Ella 87
The age of the root affects aerenchyma formation. An earlier view
was that in cereals such as com and wheat/ aerenchyma would be
formed in nodal but not seminal roots. Konings (1982) and Thomson et
al. (1990) showed that the seminal roots of corn and wheat also form
aerenchyma, provided the roots are exposed to stagnant conditions or
ethylene when they are still shorter than 1 0 - 2 0 cm. In wheat, nodal roots
longer than 1 0 cm do not develop a large volume of aerenchyma even
when exposed to conditions which induce aerenchyma formation
(Thomson et al,, 1990).
The formation of aerenchyma during flooding is not only confined
to the roots but is also seen in the leaves (Jackson, 1989, Grinieva and
Bragina, 1993 as cited by Vartapetian and Jackson, 1997) and thus
presumably interconnects with the root aerenchyma to form a
continuum (Vartapetian and Jackson, 1997).
2. S t e m e l o n g a t io n
Adaptation to submergence is a complex process involving a number of
traits. The suitable combination of these traits varies depending on the
type of submergence, i.e., flash flood or stagnant, and is generally
observed not only in rice and similar monocots, but also in some dicots,
e.g. the genus Rumex, a native of the floodplains of European rivers (as
reviewed by Blom et al,, 1990; Voesenek et al., 1992; Laan and Blpm,
1990). Like rice, Rumex consists of a range of genotypes with varying
tolerance and adaptation to submergence. Among these are Rumex
acetosa, intolerant to submergence and ethylenednsensitive so that
elongation of the petioles does not occur either during submergence or
exposure to ethylene, and growing on the high ground of river banks;
Rumex crispus, tolerant to complete submergence even in darkness with
relatively little elongation during submergence, and adapted in areas of
transient submergence; and Rumex maritimus, less tolerant to prolonged
complete submergence in darkness than Rumex crispus, with rapid
elongation of petioles during submergence and exposure to ethylene,
and like deepwater rice adapted in areas prone to prolonged
submergence. Elongation growth may compete with maintenance
processes for energy and could thereby reduce survival during
submergence. This correlative response between intolerance to complete
submergence in darkness and substantial ability to elongate was further
explored by Setter and Laureles (1996) who demonstrated with five rice
genotypes that there is a good negative correlation (r = - 0.81) between
survival and elongation growth during complete submergence. Increase
in survival with the addition of a gibberellin biosynthesis inhibitor is
most likely related to reduced elongation growth since (i) addition of
gibberellin had the opposite effect by reducing survival and (ii)
gibberellin and its biosynthesis inhibitor together had no effect on

88 Rice Breeding and Genetics: Research Priorities and Challenges
elongation growth and survival. Moreover, a gibberellin-deficient
mutant of rice with equal initial dry weights and carbohydrate levels
relative to a submergence-tolerant cultivar exhibited little elongation
during submergence and showed a high level of submergence tolerance.
This finding suggests that by selecting for less elongating genotypes,
there could be a potential to increase submergence tolerance in periods
of complete submergence when the elongation ability is not required.
Elongation under flash-flood conditions would constitute a
disadvantage because the taller plants would tend to lodge once the
water level recedes.
Mainly characteristics of the shoot contribute to obtention of oxygen
when the usual oxygen supply route is severed by flooding and stem
elongation would be a major feature, ensuring rise of leaves above the
water level and hence the ability to photosynthesize. Rapid emergence
into the aerial environment by young stems or petioles originating from
the seed or perennating organs rooted in the substratum presumably
benefit the plant by allowing undisturbed photosynthesis, aerobic
respiration, and pollination. Avoidance of submergence through stem
elongation is typical of deepwater rice, a remarkable ecotype which can
elongate its stem by up to 25 cm a day to reach 5 m or more in length and
yield grains despite several months of deepwater stress (Vergara et al.,
1976). Leaf and internode elongation is appropriate to the deep water
and floating rice areas where water level remains > 1 0 0 cm high for
several months, nonetheless plants may completely submerged for short
period if flooding is severe.
Plants partially submerged and having the ability to elongate rapidly
increase their endogeneous ethylene content by 2 to 3 times (Metraux
and Kende, 1983). The importance of ethylene in underwater growth
Was revealed by slow growth in the presence of an inhibitor of an
ethylene precursor and reversal of the effect when the ethylene precursor
was added.
3. N o d a l r o o t f o r m a t io n
A swelling at the base of the stem, called hypertrophy, quickly develops
when this region is submerged during flooding, as reported in tomato,
sunflower, corn, various woody species (Kramer, 1951; Phillips, 1964;
Kutnetsova et al., 1981; Hook, 1984 respectively, as cited in Jackson,
1985) and floating rice (Suge, 1985). This is the result of an accelerated
lateral cell expansion and can be associated with increased intercellular
space, cell lysis to form an aerenchyma, and subsequent adventitious
rooting. There is substantial evidence (Kawase, 1974; Wample and Ried,
1979) that hypertrophy in sunflower hypocotyls waterlogged over 1-3

Osamu Ito, Gloria Cabuslay and Evangelina Ella 89
days is associated with enhanced ethylene concentration in the swelling
tissue and increased cellulase activity that may soften cell walls, thirdly
favoring expansion (Kawase 1979). The same response was induced
when a water jacket was applied only to the base of hypocotyls (Kawase,
1979) or the roots were submerged with hypocotyls in aerated water
(Wample and Reid, 1975, 1978). Hypertrophy was also observed after
applying ethephon, a substance that breaks down to form ethylene in the
plant tissue (Kawase, 1974) or 0.1 Pa ethylene (Kawase, .1981). These
findings suggest that submergence alone is sufficient, anoxic roots are
not required to induce hypertrophy, and the entrapped ethylene is likely
to be the cause even though Kawase (1981) failed to show evidence that
silver ions inhibited cellulase activity and hypertrophy in submerged
hypocotyls.
B. Plant Metabolism
The effect of limited gas diffusion during submergence is demonstrated
by the observation that when a rice plant is flushed with air of high CO2
partial pressure (1-2 kPa) while being submerged, it survives for up to a
3-month complete submergence in contrast to 0.03 kPa in which the rice
plant dies within 1-2 weeks (Setter et al., 1989). The beneficial effects of
high C0 2 -flushing treatments are rather complex because such
treatments would result not only in an increase in CO2 supply, but also
in O2 and possibly in decreased concentrations or effects of ethylene ini
floodwater (Setter et al, 1997).
The effects on growth due to O2 deficiency are described by Drew
(1983), Jackson and Drew (1984), Drew (1990), Setter and Ella (1994),
Setter et al (1994), and Setter et al (1997). Anoxia for 24 h resulted in
death of many plants including an anoxia-intolerant rice cultivar IRS
(Crawford 1989) because the absence of oxygen may stop respiration
and reduce energy production for survival and elongation growth. It
inhibits oxidative phosphorylation reducing the rates of ATP synthesis
by 12- to 18-fold unless there are metabolic adaptations.
The energy shortage caused by anoxia may lead to reduced nutrient
uptake, cessation of root elongation in submerged plants, and eventually
pronounced root injury and death of the tips and the whole root systems
(Waters et al, 1989). Such effects on the root function and growth would
cause reductions in shoot growth due to nutrient deficiencies (Drew,
1983) and feedback control on the rate of shoot growth to limit excessive
increase in shoot-to-root ratio which might be mediated by plant
hormones such as ethylene Qackson and Drew, 1984).
Carbon assimilation during submergence is affected by several
factors including CO2 supply, irradiance, and the capacity of the plant to

90 Rice Breeding and Genetics; Research Priorities and Challenges
photosynthesize under floodwater. The last factor may be mediated by
ethylene effects on chlorophyll Qackson et al.^ 1987). Carbohydrate is an
important factor in conferring submergence tolerance in rice (Setter et
al.f 1997). Several researches support this conclusion based on
observations in (i) dark and shade treatments before and during
submergence (Palada and Vergara^ 1972); (ii) CO2 supply and hence
floodwater pH which affects the ratio of carbon dioxide to bicarbonate
and plant photosynthesis under water (Setter et at, 1989); (iii) seed size
(Ella and Setter, 1996; IRRI, 1996); and (iv) seedling age (Reddy and
Mittra, 1985; Chaturvedi et at, 1995; Mallik et at, 1995). There was a
strong correlation between stem carbohydrate prior to submergence
and tolerance to 12-d submergence (r^ = 0.90,30-d-old rice plants; Mallik
et at, 1995).
VI. CONCLUDING REMARKS
Although productivity is low, there is an increasing tendency for farmers
to use rainfed areas for rice cultivation. This is because of rapid
conversion of prime agricultural lands to housing and industrial uses,
because landowners receive greater returns on their investments.
Scientists, should therefore, work harder in developing cultivars
adapted to drought/submergence-prone areas to meet the demand of
subsistence farmers and the ever-increasing world population,
especially in developing countries.
Despite concerted scientific efforts, development of high-yielding
and drought submergence-tolerant rice cultivars remains elusive. This
demonstrates the complexity of rainfed ecosystems, worsened by
unpredictable moisture supply. A deeper understanding of the
mechanisms of drought/submergence tolerances in rice is necessary for
breeders to be able to identify heritable traits that will make plants
adaptable to harsh conditions in rainfed areas.
Ap Reses, T. and Wilson, P.M. 1984. Effects of reduced supply of oxygen on the metabolism
to roots of Glyceria maxima and Pisum saiivum. Z. Pflanzenphysiot 114:493-503.
Armstrong, W. 1979. Aeration in higher plants. In: Advances in Botanical [^search. H.W.W.
Woolhouse, (ed.). Acad. Press, London, pp. 225-332.
Armstrong, W., Armstrong, J. and Becket, P.M. 1996. Pressurised aeration in wetland
macrophytes: some theoretical aspects of humidity-induced convection and thermal
transpiration. Folia Geobot. Phytotax. 31:25-36.
Atkin, O.K., Villar, R. and Lambers, H. 1995. Partitioning of electrons between the
cytochrome and alternative pathways in intact roots. Plant Physio/. 108:1179-1183.

Osamu Ito, Gloria Cabuslay and Evangelina Ella 91
Atwell, B.J., Drew, M,C. and Jackson, M.B. 1988. The influence of oxygen deficiency on
ethylene synthesis, l-aminocyclopropane-l-carboxylic acid levels, and aerenchyma
formation in roots of Zea mays. Physiologia Plantamm 72; 15-22.
Bahr, J.T. and Bonner, W.D. 1973. Cyanide-insensitive respiration, II. Control of the
alternative pathway. /. Boi7. Chem. 218:3446-^450.
Baluska, F., Brailsford, R.W., Hauskrecht, M., Jackson, M.B. and Barlow, P.W. 1993.
Differential morphogenesis in the maize root cortex: involvement of microtubules and
phytohormones during post-mitotic cell growth in relation to aerenchyma and other
responses to environmental stresses. Botánica Acta 106:394-404.
Barber, D.A., Ebert, M. and Evans, N.T.S. 1962. The movement of through barley and
rice plants. J. Exp. Bot. 13: 397-403.
Begg, J.E, 1980. Morphological adaptations of leaves to water stress. In: Adaptation of Plants to
Water and High Temperature Stress. N.C. Turner and P.J. Kramer (eds.), Wiley-Interscience,
New York, N. Y„ pp. 33-42.
Boyer, J.S. 1996. Advances in drought tolerance in plants. Adv. Agron. 56:187-218.
Biom, C.W.P.M., Eogemann, G.M., harm, P., van der Sman, A.J.M., van de Steeg, H.M. and
Voesenek L.A.C.J. 1990. Adaptations to flooding in plants from river areas. Aquatic
Botany 38: 29-47.
Blom, C.W.P.M., Voesenek, LA.C.J., Banga, M., Engelaar, W.M.H.G., Rijnders,J.H.G.M., Van
Steeg H.M. and Visser E.J.W. 1994, Physiological ecology of riverside species: Adaptive
responses of plants to submergence. Ann. Botany 74:253-263.
Boyer, J. and McPherson, H.G. 1976. Physiology of water deficits in cereal grains. Proc.
Symp. Climate and Rice, IRRI, Los Banos, Laguna, Philippines, pp. 321-339.
Bradford, K.J. and Dilley, D.R. 1978. Effects of root anaerobiosis on ethylene production,
epinasty, and growth of tomato plants. Plant Physiol. 61:506-509.
Brailsford, R.W., Voesenek, L.A.C.J., Blom, C.W.P.M., Smith, A.R., Hall, M.A. and Jackson,
M.B. 1993. Enhanced ethylene production by primary roots of Ze

92 Rice Breeding and Genetics: Research Priorities and Challenges
Dingkuhn, M,, De Datta^ S.K., Dorffling, K. and Javellana, C. 1989. Varietal differences in leaf
water potential^ leaf net CO2 assimilation, conductivity and water use efficiency in
upland rice. Aust, J. Agrie. Res. 40; 1183-1192.
Dingkuhn, M., Farquhar, G.D., De Datta, S.K. and O'Toole, J.C. 1991. Discrimination of
among upland rices having different water use efficiencies. Aust. }. Agrie Res. 42:1123-
1131.
Dingkuhn, M. and Kropff, M. 1996. Growth analysis and photoassimilate partitioning. In:
Photoassimilate D/stríÍJwfi'o« in Plants and Crops. Source-Sink Relationship. E. Zamski and A.
Schaffer, (eds.). MarceJ Dekker, New York, pp. 519-547.
Drew, M.C. 1983. Plant injury and adaptation to oxygen in the root environment; a review.
Plant and Soil 75:179-199.
Drew, M.C. 1990. Sensing soil oxygen. Plant Cell Environ. 13: 681-693.
Drew, M.C., Jackson, M.B. and Giffard, S. 1979. Ethylene-promoted adventitious rooting and
development of cortical air spaces (aerenchyma) in roots may be adaptive responses to
flooding in Zea mays L. Planta 147: 83-88.
Dry, I.B., Moore, A.L., Day, D.A. and Wiskich, J.T. 1989. Regulation of alternative pathway
activity in plant mitochondria. Non-linear relationship between electron flux and the
redox poise of the quinone pool. Arch. Biochem. Biophys. 273; 148-157.
Eghball, B. and Maranville, J.W. 1993. Root development and' nitrogen influx of corn
genotypes grown under combined drought and nitrogen stresses. Agron. J. 85; 147-152.
Ella, E.S. and Setter, T.L. 1996. Importance of carbohydrates in rice {Oryza sativa L.) seeds
exposed to anoxia. Proc. 12th Ann. Sei. Conf. Fed. Crop Sei. Soci. Philippines, Davao City,
Philippines, 21 suppl. 1: 75.
Farquhar, G.D. and Richards, R.A. 1984. Isotopic composition of plant carbon correlates
with water-use efficiency of wheat genotypes. Aust f. Plant Physiol. 11:539-552.
Fukai, S. and Cooper, M. 1995. Development of drought-resistant cultivars using
physiomorphological traits in rice. Field Crops Res. 40: 67-^6.
Gambrell, R.P., Delaune, R.D., and Patrick, W.H. Jr. 1991. Redox processes in soils following
oxygen depletion. In: Plant Life Under oxygen deprivation. Ecology, physiology, and
biochemistry. M.B. Jackson, D.p, Davies and H. Lumbers (eds). The Hague. SPB Academic,
pp. 101-117.
Grinieva, G.M. and Bragina, T.V. 1993. Structural and functional patterns of maize
adaptation to flooding, Russian Plant Physiol. 40:662-667.
Hanson, A.D., Peacock W.J., Evans L.T., Amtzen C.J., and Khush, G.S. 1995. Drought
resistance in rice. Nature 234; 2, In: Development of drought-resistant cultivars using
physio-morphological traits in rice. Fukai, S. and Cooper M., (ed.). Field Crops Res. 40;
67-86.
He, C.J., Drew, M.C. and Morgan, P.W. 1994. Induction of enzymes associated with
lysigenous aerenchyma formation in roots of Zea mays during hypoxia or nitrogen
starvation. Plant Physiol, 105; 861-865,
He, J.X., Wang, J. and Liang, H.G. 1995. Effects of water stress on photochemical function and
protein metabolism of photosystem II in wheat leaves. Physiol. Plant. 93: 771-777.
Heckman, C.W. 1979. Ricefield ecology in Northeastern Thailand: the Effect of Wef and Dry
Seasons on a Cultivated Aquatic Ecosystem. Dr. W. Junk (ed.) Publishers, London.
Fluang, B. and Johnson, J.W. 1995. Root respiration and carbohydrate status of two wheat
genotypes in response to hypoxia. Ann. Botany 75:427-432,
Huang, B., Johnson, J.W., Nesmith, S, and Bridges, D.C. 1994, Growth, physiological, and
anatomical responses of two wheat genotypes to waterlogging and nutrient supply. J.
Exp. Bot 45; 193-202.
International Rice Research Institute (IRRI) 1990. Program Report for 1989. Los Baños,
Laguna, Philippines, p. 94.

Osamu lio, Gloria Cabuslay and Evangelina Ella 93
IRRI. 1996. Program Report for 1995. Los Banso, Laguna, Philippines, pp. 88,92.
Ito, O., Katayama, K., Adu-Gyamfi, J.J., Devi, G. and Rao, T.P. 1996. Root activities and
function in component crops for intercropping. In: Roots and Nitrogen in Cropping Systems
of the Semi-Arid Tropics. O. Ito, C. Johansen, J.J. Adu-Gyamfi, K. Katayama, J.V.D.K.
Kumar Rao and T.J. Rego, (eds.). Proc. Inti. Workshop Dynamics of Root and Nitrogen in
Cropping Systems of the Semi-Arid Tropics (ICRISAT) at Patancheru, India, 21-25
November 1994, pp. 149-172.
Jackson, M.B. 1979. Is the diageotropic tomato ethylene-deficient? Pftysioi. Plant 46:347-351.
Jackson, M.B. 1985. Ethylene and responses of plants to soil waterlogging and submergence.
Ann. Rev. Plant Physiol. 36:145-174.
Jackson, M.B. 1989. Regulation of aerenchyma formation in roots and shoots by oxygen and
ethylene. In: Cell Separation in Plants: Physiology, Biochemistry, and Molecular Biology. D.J.
Osborne and M.B. Jackson, (eds,). Springer Berlin, pp. 263-274.
Jackson, M.B. and Attwood, P. 1996. Roots of willow {Salix viminalis L.) show marked
tolerance to oxygen shortage in flooded soils and solution culture. P/

94 Rice Breeding and Genetics: Research Priorities and Challenges
Laan, P., Berrevoets M.J., Lythe S., Armstrong, W., and Blom, CWPM. 1989. Root morphology
and aerenchyma formation as indicators for the flood-tolerance of Rumex species. J, Eco/.
77; 693-703.
Lambers, H., van der Wèrf, A, and Konings, H. 1991. Respiratory patterns in roots in relation
to their functioning. In; Plant Roots: The Hidden Half. Waisel, Eshel and Kafkafi, (eds.), pp.
229-263.
Laude, H.M. 1971. Drought influence on physiological processes and subsequent growth. In:
Drought Injury and Resistance in Crops. Proc. Symp. Ann. Meeting Crop Sei. Soci. Amer,
and Amer. Seo. Agron., Tucson, Arizona, August 24,1970.
Lawlor, D.W. 1995. The effects of water deficit on photosynthesis. In; Environment and Plant
meíflí?o//sm. F/ej:ií)27ifi/ and Acclimation. N. Smirnoff, (ed.) BIOS Sei. Publ. UK,pp. 129-160.
Levitt, J. 1980, Responses of Plants to Environmental Stresses, Vol. II. Water, Radiation, Salt, and
Other Stresses. Acad. Press, NY.
Lilley, J.M. and Fukai, S. 1994 a. Effect of timing and severity of water deficit on four diverse
rice cultivars, I. Rooting pattern and soil water extraction. Field Crops Res. 37(3): 205-213.
Lilley, Ï.M. and Fukai, S. 1994b. Effect of timing and severity of water deficit on four diverse
rice cultivars, III. Phenological development, crop growth and grain yield. Field Crops
Res. 37; 225-234.
Luxmoore, R.Ï. and Stolzy, L.H. 1969. Root porosity and growth response of rice and maize
to oxygen supply. Agron. J. 61:202-204.
Ma, y.F. and Lu, D.Z. 1990. Effect of irrigation methods on senescence and physiological
activities in hybrid rice after heading. Chinese /. Rice Sei, 4; 56-62.
Mackill, D.J., Coliman, W.R. and Garrity, D.P. 1996. Rainfed lowland rice improvement.
ÏRRI, Manila, Philippines.
Mallik, S., Kundu, C., Baneiji, C., Nayak, D.K., Chaterjee, S.D., Nanda, P.K., Ingram, K.T. and
Setter, T.L., 1995. Rice germplasm evaluation and improvement for stagnant flooding. In:
Rainfed Lowland Rice-Agricultural Research for High Risk Environment. K.T. Ingram, (ed.)
IRRI, Manila, pp, 97-109.
Metraux, J.P. and Kende, H. 1983. The role of ethylene in the growth response of submerged
deep water rice. Plant Physiol. 72: 441-446.
Millar, A.H., Atkin, O.K., Lambers, H., Wiskich, J.T. and Day, D. A. 1995. A critique of the use
of inhibitors to estimate partitioning of electrons between mitochondrial respiratory
pathways in plants. Physiol, Plant. 95:523-532.
Moreshet, S., Bridges, D.C., NeSmith, D.S. and Huang, B. 1996. Effects of water deficit stress
on competitive interaction of peanut and sicklepod. Agron. J. 8 8 ; 636-651.
Morgan, J.M. 1984. Osmoregulation and water stress in higher plants. A««. Reí?. Plant Physiol.
35: 299-319.
Morita, S. and Abe, J. 1996. Development of root systems in wheat and rice. In: Roofs and
Nitrogen in Cropping Systems of the Semi-Arid Tropics. O. Ito, C. Johansen, J.J. Adu-Gyamfi,
K, Katayama, J.V.D.K. Kumar Rao and T.J. Rego, (eds.). Proc. Inti. Workshop Dynamics
of Root and Nitrogen in Cropping Systems of the Semi-Arid tropics (ICRISAT) at
Patancheru, India, 21-25 November 1994. pp. 185-198,
Nicolas, M.,Lambers, E.H., Simpson, R.J., and Dalling, Wf.J. 1985. Effect of drought on
metabolism and partitioning of carbon in two varieties of wheat differing in drought
tolerance, Ann, Bot. 55: 727-742,
O'Toole, J.C. and Chang, T.T. 1978. Drought and rice improvement in perspective, IRRI
Research Paper Series (14): 1-27.
Palada, M. and Vergara, B.S. 1972. Environmental effect on the resistance of rice seedlings to
complete submergence. Crop Sei. 12:209-212.

Osamu ItO/ Gloria Cabuslay and Evangelina Ella 95
Pardales, J.R. Jr. and Kono, Y. 1990. Development of sorghum root system under increasing
drought stress, fap. f. Crop Sci. 59(4): 752-761.
Poorter, H., van der Werf, A., Atkin, O.K. and Lambers, H. 1991. Respiratory energy
requirements of roots vary with the potential growth rate of a plant species. Physiol.
Plant. 83:469-475.
Ram, P.C., Singh, B.B., Singh, A.K., Singh, V.K. and Singh, V.P. 1996. Environmental and
plant measurement requirements for the assessment of drought, flooding, and salinity
tolerance in rice. In: Proc. Inti. Conf, Stress Physiology of Rice, 28 Feb.-5 Mar. 1994,
Lucknow, India, pp. 44-69.
Reddy, M,D. and Mittra, B.N. 1985. Effects of complete plant submergence on vegetative
growth, grain yield, and some biochemical changes in rice plants. Plant and Soil 87:365-
374,
Saini, H.S. 1997. Effects of water stress on male gametophyte development in plants. Sex
Plant Reprod. 10: 67-73.
Samson, B.K., Wade, L.J., Sarkarung, S., Hasan, M. Amin, R., Hampichitvitya, D., Pantuwan,
G., Rodriguez, R., Sigari, T. and Calendacion, A, 1995. Examining genotypic variation in
root traits for drought resistance. In: Fragile Lives in Fragile Ecosystem. Proc. Rice Research
Conf., 13-17 February, 1995. IRRI, Los Banos, Laguna, Philippines, pp. 521-534.
Sarkarung, S., Singh, O.N., Roy, J.K., Vanavichit, A. and Bhekasut, P. 1995. Breeding
strategies for rainfed lowland ecosystem. In: Fragile Lives in Fragile Ecosystems. Proc. Inti.
Rice Research Conf. 13-17 February, 1995, IRRI, Los Banos, Laguna, Philippines, pp. 709-
720.
Setter, T.L. and Ella, E.S. 1994. Relationship between coleoplile elongation and alcoholic
fermentation in rice exposed to anoxia, I. Importance of treatment conditions and
different tissues. Ann. Botany 74:265-271.
Setter, T.L. and Laureles, E.V. 1996. The beneficial effect of reduced elongation growth on
submergence tolerance in rice. /. Exp. Bot. 47:1551-1559.
Setter, T.L., Ella, E.S. and Valdez, A.P. 1994, Relationship between coleoptile elongation and
alcoholic fermentation in rice exposed to anoxia, II. Cultivar difference Anna. Botany 74:
273-279.
Setter, T.L., Ellis, M., Laureles,, E.V., Ella, E.S., Senadhira, D., Mishra, S.B., Sarkarung, S, and
Datta, S, 1997. Physiology and genetics of submergence tolerance in rice. Ann. Botany 79
(Suppl. A): 67-77.
Setter, T.L, Ingram, K.T. and Tuong, T.P. 1995. Environmental characterisation requirements
for strategic research in rice grown under adverse conditions of drought, flooding, or
salinity. In: Rainfed Lowland Rice-Agricultural Research for High-Risk Environments, K.T.
Ingram, (ed.), IRRI, Manila, pp. 3-18.
Setter, T.L., Kupkanchanakul, T,, Kupkanchanakul, K., Bhekasut, P,, Wicngweera, A. and
Greenway, H. 1987. Concentrations of COj and O2 in floodwater and in internodal
lacunae of floating rice growing at 1-2 meter depths. Plant Cell Environ. 10: 7&7-T/fi.
Setter, T.L,, Kupkanchanakul, T., Waters, I. and Greenway, H. 1988a. Evaluation of factors
contributing to diurnal changes in floodwater in deep water rice fields. New Phyiologist
110: 151-162.
Setter, T.L., Kupkanchanakul, T., Kupkanchanakul, K., Bhekasut, P., Wiengweera, A. and
Greenway, H. 1988b, Environmental factors in deep water rice areas in Thailand: O^,
CO2, and ethylene. In: Symp. 1987 Intel. Deepwater Rice Workshop. IRRI, Los Baños,
Philippines, pp. 55-66.
Setter, T.L., Waters. L, Wallace, I., Bhekasut, P. and Greenway, H. 1989. Submergence of rice,
I. Growth and photosynthetic response to CO2 enrichment of floodwater. Aust. J. Plant
Physiol 16: 251-263,

96 Rice Breeding and Genetics: Research Priorities and Challenges
Shih, S.F., Rahi, G.S. and Harrison, D.S., 1982. Evapotranspiration studies on rice in relation
to water use efficiency. Trans. ASAE 25(3); 702-707.
Singh, A.K., Tuong, T.P., Woperies, M.C.S., Boling, A. and Kropff, M.J. 1995. Quantifying
lowland rice responses to soil-water deficit. In: Fragile Lives in Fragile Ecosystems. Proc.
Intel. Rice Research Conf. 13-17 February, 1995, IRRI, Los Banos, Laguna, Philippines,
pp. 507-519.
Singh, H. and Ingram, K.T. 1991. Sensitivity of rice to water deficit at different growth stages.
Phil.}, Crop Sei. 16 (SuppL No. 1); S ll.
Sinhababu, D.P., Panda, M.M. and Rajamani, S., 1991. Compatibility of dhaincha (Sesbania
aculeata) with rice-fish in rainfed intermediate lowlands (0-^0 cm). Onjza^28:497-A99.
Slatyer, R.O. 1973. The effect of internal water status on plant growth, development and
yield. In; Plant Response to Climatic Factors. R.O. Slatyer, (ed.). Proc. Uppsala Symp.
UNESCO, Paris, pp. 177-191.
Smirnoff, N. and Crawford, R.M.M. 1983. Variation in the structure and response to flooding
of root aerenchyma in some wetland plants. Ann. Botany 51:237-249.
Smith, F.A. and Walker, N.A. 1980. Photosynthesis by aquatic plants—^Effects of unstirred
layers in relation to assimilation of CO2 and HCO3 and to carbon isotopic
discriminationn. Neio Phytologist 8 6 : 245-259.
Suge, H. 1985. Ethylene and gibberellin. Regulation of internodal elongation and nodal root
development in floating rice. Plant Cell Physiol. 26:607.
Taiz, L. and Zeiger, E. 1991. Plant Physiology. Benjamin/Cummings Publ. Co., pp. 346-^56.
Thomson C.J., Armstrong, W., Waters, I., and Greenway, H. 1990. Aerenchyma formation
and associated oxygen movement in seminal and nodal roots of wheat. Plant Cell Environ.
13: 395-403.
Van den Boogaard, R., Kostadinova, S., Veneklaas, E. and Lambers, H. 1995. Association of
water use efficiency and nitrogen use efficiency with photosynthetic characteristics of
two wheat cultivars. /. Exptl. Bot. 46:1429-1438.
Van der Wafer, A. 1996. Growth analysis and photoassimilate partitioning. In:
Photoassimilate Distribution in Plants and Crops. Source-Sink Relationship. E. Zamski and A.
Schaffer (eds.). Marcel Deeker, New York, pp, 1-6.
Varade, S.B., Stolzy, L.H. and Letey, J. 1970. Influence of temperature, light intensity, and
aeration on growth and root porosity of wheat (Tritkum aestivum), Agron.}. 62: 79-83.
Vartapetian, B.B. and Jackson, M.B, 1997. Plant adaptations to anaerobic stress. Ann. Botany
79 (Suppl. A); 3y20,
Vergara, B.S., Jackson, B. and De Datta, S.K. 1976. Deep water rice and its response to deep
water stress. In; Climate and Rice. Intel. Rice Research Institute, IRRI Los Baños
Philippines, pp. 301-319.
Voesenek, L.A.C.J., van der Sman, A.J.M., Harren, F.J.M. and Blow, C.W.P.M. 1992. An
amalgamation between hormone physiology and plant ecology: a review on flooding
resistance and ethylene. /. Plant Growfii Regulation 11:171-188.
Wade, L.J., Sarkarung, S., McLaren, C.G., Guhey, A., Quader, B,, Boonwite, C,, Amarante,
S.T., Sarawgi, A.K., Haque, A., Harnpichitvitya, D., Pamplona, A. and Bhamri, M.C. 1995.
Genotype x environment interaction and selection methods for identifying improved
rainfed lowland rice genotypes. In; Fragile Lives in Fragile Ecosystems. Proc. Intel. Rice
Research Conf. 13-17 February, 1995, IRRI, Los Baños, Laguna, Philippines, pp. 885-900.
Wagner, A.M. and Krab, K. 1995, The alternative respiration pathway in plants: Role and
regulation. Physiol. Plant. 95:318-325.
Wample, R.L. and Reid, D.M. 1975, Effect of aeration on the flood-induced formation of
adventitious roots and other changes in sunflower {Heliantitus annuus L.). Planta 127; ’
263-270.

Osamu Ito, Gloria Cabuslay arvd Evangelina Ella 97
Wample, R.L. and Reid, D.M. 1978. Control of adventitious root production and hypocotylhypertrophy
of sunflower (Helianthus annuus L.) in response to flooding. Physiol. Plant.
44: 351-358.
Wample^ R,L. and Reid, D.M. 1979. The role of endogeneous auxins and ethylene in the
formation of adventitious roots and hypocotyl hypertrophy in flooded sunflower plants
(Helianthus annus L.). Physiol. Plant. 45:219-226.
Wardlaw, I,F. 1990. The control of carbon partitioning in plants. New Phytol. 116: 341-381.
Waters, I., Armstrong, W., Thomson, C.J., Setter, T.L., Adkins, S., Gibbs, J. and Greenway, H,
1989. Diurnal changes in radial oxygen loss and ethanol metabolism in roots of
submerged and non submerged rice seedlings. New Phytologist 113; 439-451.
Webb, J. and Jackson, M.B. 1986. Transmission and cryo-scanning electron microscopy study
of the formation of aerenchyma (cortical gas-filled space) in adventitious roots of rice
(Oryza sativa L.) /. Exp. Bot^3>7: 832-^41.
Whitton, B.A., Aziz, A., Francis, P., Rother, J.A., Simon, J.W. and Tahmida, 2.N. 1988.
Ecology of deepwater rice fields in Bangladesh, I. Physical and chemical environment.
Hydrobiologia 169; 3-67.
Wonprasaid, S., Khunthasuvon, S., Sittisuang, P. and Fukai, S. 1996. Performance of
contrasting rice cultivars selected for rainfed lowland conditions in relation to soil
fertility and water availability. Field Crops Res. 47:267-275.
Yamauchi, A., Pardales, J.R.Jr., and Kono, Y. 1996, Root system structure and its relation to
stress tolerance. In: Roofs and Nitrogen in Cropping Systems of the Semi-Arid Tropics. O. Ito,
C. Johansen, J.J. Adu-Gyamfi, K. Katayama, J.V.D.K. Kumar Rao and T.J. Rego (eds.).
Proc. Inti. Workshop D^am ics of Root and Nitrogen in Cropping Systems of the Semi-
Arid Tropics (ICRISAT) at Patancheru, India, 21-25 November 1994, pp. 211-233,
Yamasaki, T. 1952. Studies of the 'excess moisture injury' of upland crops in overmoist soil
from the viewpoint of soil chemistry and plant physiology. Bull. Nat Inst. Agri. Sei.
(Japan). B. 1:1-92.
Yu, P.T., Stolyz, L.H. and Letey, J. 1969. Survival of plants xmder prolonged flooded
conditions. Agron. J. 61:844-847,

New Plant Type of Rice for
Increasing the Genetic Yield
Potential
Gurdev S, Khush’^
INTRODUCTION
Major increases in rice production have occurred during the last 25
years because of large-scale adoption of high-yielding semidwarf
varieties and improved management practices. World rice production
doubled in a 25-year period/ from 256 Mt in 1966 to 520 Mt in 1990.
During this period/ rice production increased at a slightly higher rate
than the population. However/ the rate of increase of rice production is
now lower (1.5% per year) than the rate of increase of population (1.8%
per year). If this trend is not reversed/ severe food shortage will occur in
the next century. The present world population of 5.8 billion is likely to
reach 6.0 billion in 2000 and 8.2 billion in 2025. The population of rice
consumers is rapidly increasing and will probably increase by 60% in the
next 25 years. It is estimated that demand for rice will exceed production
by the early part of the next century (Pinstrup-Anderson/ 1994).
Major increases in the area planted to rice are unlikely. The area
planted to rice has remained stable since 1980. Moreover/ a diminution in
the area planted to rice is likely because of the pressures on good rice land
from urbanization and industrialization. The increased demand for rice *
* International Rice Research Institute, P.O. Box 933, Manila, Philippines.

100 Rice Breeding and Genetics: Research Priorities and Challenges
will have to be met from less land, with less water, less labor, and less
pesticides. Therefore, rice varieties with higher yield potential and better
management practices are needed to meet the goals of increased rice
production. In its strategy document for 2000 and beyond, the
International Rice Research Institute (IRRI) accorded highest priority to
increase the genetic yield potential of rice (IRRI, 1989).
The yield potential of current high-yielding rice varieties in the
tropics is 10 t ha'^ during the dry season and 6.5 t ha'^ during the wet
season. Plant physiologists have suggested that physical enviroranent in
the tropics is not a limiting factor .for increasing rice yields. Maximum
yield potential was estimated to be 9.5 t ha'^ during the wet season and
15.9 t ha'^ during the dry season (Yoshida, 1981).
MODIFICATION OF RICE PLANT IN THE 1960s
Pre-green revolution rice varieties were tall and leafy with weak stems
and produced a total biomass (leaves, stems, and grain) of about 12 t.
When nitrogenous fertilizer was applied at rates exceeding 40 kgha'^,
these traditional varieties tillered profusely, grew excessively tall, lodged
early, and the biomass could not be increased by fertilization. These
varieties had a harvest index (ratio of dry grain weight to total dry matter
or biomass) of 0.3. Thus of 12 tons biomass, about 30% was grain or a
maximum yield of about 4 t ha"\ To increase the yield potential of
tropical rice, it was necessary to improve the harvest index as well as
total biomass or nitrogen responsiveness. This was accomplished by
reducing the plant height through the incorporation of a recessive gene
sd~l for short stature from a Chinese variety Dee-geo-woo-gen (Fig. 5.1)
(Khush, 1995). The first semidwarf variety IR8, developed at IRRI, also
had a combination of other desirable features such as profuse tillering,
dark green and erect leaves, and sturdy stems. It had a harvest index of
0.45 and did not lodge when high doses of nitrogenous fertilizer were
applied. With proper fertilization it could produce 18 t biomass per ha
and a grain yield of 8 1 ha'\ Thus the yield potential of tropical rice was
doubled from 4 to 8 t ha'^ through modification of plant type. Since the
development of IR8 in 1966, the yield potential of rice has been improved
at the rate of about 1% per year. Thus IR72 released in 1980 produces a
biomass of about 20 t ha'^ and a harvest index of 0.5. It yields
10 t ha'^ under proper management.
NEW PLANT TYPE FOR INCREASED YIELD POTENTIAL
As discussed above, yield is a function of total dry matter or biomass
and the harvest index. Therefore, to further increase the yield potential

Gurdev S. Khush 101
Fig. 5.1
Plots of a conventional rice variety (left) and an improved high-yielding
variety (right).
of tropical rice^ we have to either increase the total biomass production
or the harvest index or both. We conceptualized a plant type to increase
the biomass to about 23 and harvest index to 0.55. Such a plant should
produce a grain yield of about 12.5 t or an increase of 25% over the yield
of existing high-yielding varieties*
The harvest index can be increased by increasing the proportion of
energy stored in the grain or by increasing the sink size. The sink size
can be increased by:
• larger number of spikelets per panicle or ear,
• greater partition of photosynthesis in spikelet formation,
• increased spikelet filling,
• slow leaf senescence,
• maintenance of healthy root system, and
• increased lodging resistance
Biomass can be increased by both genetic manipulation as well as by
better management practices. Varietal characteristics for increasing the
biomass include;
• establishment of desirable leaf canopy structure,
• rapid leaf area development,
• rapid nutrient uptake, and
• increased lodging resistance
To achieve these goals, a new plant type was conceptualized with the
following attributes (Fig. 5.2) (Khush, 1994).
• lower tillering capacity,
• no unproductive tillers.

102 Rice Breeding and Genetics: Research Priorities and Challenges
- . M l
'V /-''’M '-vt .SM' i ■ '. : ■
r !/
/f'íMvV
\hAA/X
' ■
■'"MV'-T';
■ A y
'/ 0 V
/fw-"
' ¥ :/m A
■ \ a
Fig. 5.2
Sketches of different plant types of rice. Left: tall traditional plant type;
Center; improved high-yielding high tillering plant type; Right:
proposed low tillering ideotype of super rice.
I; i
• 200-250 grains per panicle^
• 90-100 cm height,
• very sturdy stems,
• dark green thick and erect leaves,
• thicker and deeper roots,
• multiple disease and insect resistance, and
• acceptable grain quality
Reduced Tillering
Increases in the yield potential of other cereals such as maize and
sorghum have resulted from increases in sink size (ear size). Selection
and breeding for large sink size were accompanied by a decrease in tiller
number. Modern maize and sorghum varieties have a single culm (stem),
whereas primitive maize and sorghum had a large number of tillers with
small cobs (or ears) (Khush, 1993). Teosinte, the ancestor of maize, has
20-25 tillers and small cobs with a few grains. The agriculturists who
domesticated maize in the Americas continued to select maize with large
cobs and this resulted in reduced tiller number. By the fifteenth century,
when maize was introduced into Europe, it had only 4-5 tillers. Further

Gurdev S, Khush 103
selection resulted in uniculm plants y/ith very large cobs. Single culm
sorghums were bred in the post-Mendelian era.
By contrast, modern rice varieties produce 20-25 tillers under
favorable growth conditions. Only 14-15 of these tillers produce
panicles, w h ich , are small, and the rest remain unproductive.
Unproductive tillers compete with productive tillers for solar energy
and mineral nutrients— especially nitrogen. Elimination of the
unproductive tillers could direct more nutrients to grain production.
Furthermore, dense foliage that results from excess tiller production
creates a humid microenvironment favorable to disease and insect buildup.
Reduced tillering also facilitates synchronous flowering and
maturity, and more uniform panicle size. Genotypes with lower tiller
number are also reported to produce a larger proportion of heavier
grains (Padmaja Rao, 1987).
The number of grains produced per unit of land area primarily
determines the grain yield in cereal crops grown in high-yield
environments without stress (Takeda, 1984). Increasing the number of
panicles or the number of grains per panicle can increase the number of
grains per unit land area. The modern high-yielding varieties have a
much higher panicle number than the traditional rice varieties they have
replaced. There is a limit to how much the number of panicles can be
increased. Additional tillers become unproductive, produce excessive
vegetative growth, and have a higher proportion of unfilled grains.
Hence, the approach is to increase the number of grains per panicle rather
than the number of panicles per imit area.
Grain Density and Grain-filling Percentage
Larger grains tend to be chalky and thus have a lower market value.
Medium-size grains with high density (higher specific gravity) are more
desirable. High-density grains tend to occur on primary branches of. the
panicle, whereas the grains of the secondary branches have lower density
(Ahn, 1986). Low tillering varieties have more primary tillers per unit
land area. Thus, they should produce a higher proportion of high-density
grains and contribute to increased yield potential,
A variable proportion of grains in rice varieties remains unfilled.
Higher grain-filling rates are essential to achieve maximum yield.
Photosynthate (carbohydrate) production in leaves and stems, and its
translocation and accumulation in the grains, are prerequisites for higher
grain-filling rates. For higher photosynthate production, dark green and
thick leaves that senesce (die) slowly are desirable. Thicker stems have
more vascular bundles, which provide a more effective system to
transport photosynthates to the grains. Thicker stems also contribute to
lodging resistance.

104 Rice Breeding and Genetics: Research Priorities and Chgillenges
Leaf Characteristics
Light is used more efficiently if the leaves are erect (Yoshida, 1976).
Therefore an erect leaf angle is a desirable trait for achieving high yield.
Photosynthesis is greater when leaves are exposed to light on both sides,
as in erect leaves. Droopy leaves are exposed to light only on one side.
They also raise the relative humidity and reduce the temperature since
they lessen light penetration and air movement. Thicker leaves generally
contain more nitrogen per unit leaf area and have a higher photosynthetic
rate. Thick and green leaves may show slower leaf senescence.
Growth Duration
The optimum growth duration for maximum rice yield in the tropics is
thought to be 120 days from seed to seed. In transplanted rice, varieties of
shorter growth duration usually give lower yields when planted at
conventional spacing, since they do not produce sufficient vegetative
growth for maximum yield (Yoshida, 1976). Growth duration of about
120 d allows the plant to utilize more soil, nitrogen and solar radiation,
and results in higher yields. However, for adaptation to various cropping
systems, varieties with varying growth duration of 100-130 d are
required.
Plant Height, Stem Thickness, Biomass Production
Short stature reduces the susceptibility of rice crop to lodging and leads
to a higher harvest index. A plant height of 90-100 cm is considered
ideal for maximum yield. Increased biomass production is not difficult
to achieve when the rice crop is grown in a high solar radiation
environment, provided there is an ample supply of nitrogen (Akita,
1989). Without strong, thick culms, however, increased biomass results in
lodging, mutual shading, increased disease incidence, and lower yield
(Vergara, 1988). Hence the importance of sturdy stems for lodging
resistance in raising the biomass production.
Roots are foundations of plants, yet they remain relatively unstudied
compared to the rest of the plant. Rice varieties differ as much in the plant
parts below the soil surface as in the parts above ground. For example,
different cultivars are known to have different lengths, degrees of
branching, volumes and thickness. Thicker and deeper roots provide

Gurdev S. Khush 105
better anchorage and lodging resistance. Healthy roots are more efficient
at supplying nutrients, particularly during the grain-filling period.
Disease and Insect Resistance
For the full expression of yield potential, genetic resistance to diseases
and insects is essential. Resistance is even more important under tropical
conditions. Major diseases of rice are blast, bacterial blight, sheath blight,
grassy stunt and tungro viruses. Major insects are brown planthopper,
green leafhopper and stem borers.
Grain Quality
In the tropics and subtropics, consumers prefer long, slender and
translucent grains with intermediate amylose content and intermediate
gelatinization temperature. Rices with such characteristics are soft and
moist when cooked and remain soft upon cooling. Short grained rices
with low amylose content and low gelatinization temperature are sticky
when cooked and are preferred in temperate areas such as Japan, Korea,
and China.
BREEDING FOR NEW PLANT TYPE
Breeding work on the new plant type popularly known as "super rice"
was started in 1989. About 2,000 entries from the IRRI germplasm bank
were grown to identify parents or donors for various traits. Parents for
low tillering, large panicles, thick stems, a vigorous root system, and
thick dark green leaves were identified. Hybridization was undertaken in
the 1990 dry season. To begin with, these parents were crossed with a
short statured parent and breeding lines with short stature and the
aforementioned traits were selected. These lines were intercrossed and
plant types with the proposed ideotype were selected. To date, over 2,100
crosses have been made and more than 110,000 pedigree nursery rows
have been grown. Numerous breeding lines with targeted traits have
been selected. These lines have short stature, 6-10 tillers, dark green and
erect leaves and large panicles with 200-250 grains per panicle. A plant
of one such line is shown in Fig. 5.3. The yield potential of these new
plant type lines is being evaluated in replicated yield trials under various
management practices. On the basis of observational trials the following
characteristics of the new plant type lines (NPT) have been observed:
• The NPT lines produced 6-10 tillers versus 25-27 tillers for
modern high-yielding lines.

106 Rice Breeding and Genetics: Research Priorities and Challenges
I
Fig. 5.3
Plants of super rice (left) and modern high yielding variety (right).
The number of grains per panicle in the NPT lines was 2-3 times
greater than IR72, the modern high-yielding variety.
Some of the NPT lines had 20% more grains per square meter of
land than IR72 and thus a 20% larger potential sink size.
Photosynthesis per unit area of single leaves in the new plant type
lines was 10-15% higher than that of IR72 at the vegetative and
reproductive phases. This advantage was mainly due to a higher
concentration of leaf nitrogen in the NPT lines.
The NPT lines had greener, thicker, and more erect leaves than
those of IR72. They had one or two more functional leaves at
flowering compared to IR72.
The flag leaves of NPT lines (the keys to photosynthesis during
grain filling) appear to function photosynthetically longer than
those of IR72. This may result in a longer grain-filling period and
contribute to increased yield potential.

Gurdev S. Khush 107
• The NPT lines have thicker and sturdier stems, and much greater
lodging resistance.
• The growth duration of NPT lines is similar (110-120 d) to that of
IR72.
Improvements Still Needed
l^ce germplasm is classified into two broad groups-indica and japorüca.
Indica varieties are grown in the tropics and subtropics. Modern highyielding
varieties, grown in the tropics and subtropics belong to the
indica group while those grown in the temperate areas belong to japónica
group. Some japónica varieties are grown in the tropics and are called
tropical japónicas in contrast to temperate japónicas grown in Japan,
Korea, and northern China. Improved indicas have long slender grains
whereas most of the japónicas have short bold grains. For developing the
NPT lines, tropical japónica parents were used because many of them
have large panicles, some have low tillering ability, and others have
thick stems and a vigorous root system. However, they have short bold
grains. Consequently, the NPT lines also have short bold grains.
Preference in the tropics is for long slender grains. Therefore, a few
tropical japónica parents with long slender grains were identified and
crossed with NPT lines. NPT lines with long slender grains are now
being selected.
Most of the NPT lines lack resistance to tropical diseases and insects
as the parents used for developing these lines are susceptible. However,
donors for blast and bacterial blight were identified within the tropical
japónica germplasm and utilized in the hybridization program. NPT lines
with resistance to blast and bacterial blight have now been selected.
Genes for resistance to brown planthopper, green leafhopper and tungro
viruses are being incorporated from the indica germplasm through
backcrossing and molecular marker-aided selection.
PROSPECTS FOR NEW PLANT TYPE
It is expected that during the next 3-4 years, NPT lines with all the
desirable traits will become available. These lines will be shared with the
national rice improvement programs and evaluated under local
conditions. Those found suitable will be multiplied and released for onfarm
production. Thus NPT varieties should be widely grown by 2005.
When planted widely in the tropics and subtropics, they will help feed
300 million more rice consumers.
NPT lines could not be adapted for growing in the temperate areas
as they lack cold tolerance, one of the most important adaptability traits

108 Rice Breeding and Genetics: Research Priorities and Challenges
of temperate rices. However, NPT lines could be useful parents for
increasing the yield potential of temperate rice. Collaboration has been
established wiüi the rice improvement programs of Korea and Egypt to
breed NPT varieties adapted to local conditions.
NPT varieties will also be useful for increasing the level of heterosis
(yield advantage) of hybrid rice. Most of the hybrid rices grown in China
and elsewhere are based on indica germplasm and show yield advantage
of 10-15%. The level of heterosis depends on the genetic distance
between the parents. Since the NPT lines are based on japónica
germplasm, hybrids between them and the elite indica lines show yield
advantage of 20-25% (Khush and Aquino, 1994). Thus the NPT rices will
also be useful in increasing the yield potential of hybrid rice.
References
Ahn, J.K. 1986. Physiological factors affecting grain filling in rice. Ph.D. thesis, Uiriv.
Philippines, Los Baños, Philippines.
Akita, S. 1989, Improving yield potential in tropical rice. In: Progress in Irrigated Rice Research,
(IRRI), Manila, Philippines, pp. 41-73.
IRRI (International Rice Research Institute) 1989. IRRI towards 2000 and beyond, (IRRI), Los
Baños, Philippines.
Khush, G.S. 1993. Varietal needs for different environments and breeding strategies. In: New
Frontiers in Rice Research, K. Muralidharan and E.A. Siddiq (eds.) Directorate of Rice
Research, Hyderabad, India, pp. 68-75.
Khush, G.S. 1994, Increasing the genetic yield potential of rice: prospects and approaches.
Inti. Rice Comm. News/., 43:1-8.
Khush, G.S. 1995. Modern varieties—their real contribution to food supply and equity. Geo).
35(3): 275-284.
Khush, G.S. and Aquino, R.C. 1994. Breeding tropical japónicas for hybrid rice production.
In: Hj/brid Rice Technology: New Developments and Future Prospects. S.S., Virmani, (ed.)
IRRI, Manila, Philippines, pp, 33-36.
Padmaja Rao, S. 1987. High density grain among primary and secondary tillers of short- and
long-duration rices. Inti. Rice Res. Newsl. 12(4): 12.
Pinstrup-Anderson, P. 1994. World food trends and future food security. Food Policy Report.
Inti Food Policy Res. Inst., Washington, DC.
Takeda, T. 1984. Physiological and ecological characteristics of high yielding varieties of
lowland rice. In: Proc. Inti. Crop Sci. Symp., Fukuoka, Japan.
Vergara, B.S. 1988. Raising the yield potential of rice. Philippine Tech. f. 13:3-9.
Yoshida, S. 1976. Physiological consequences of altering plant type and maturity. In: Proc.
Inti Rice Res. Conf., Inti. Rice Res. Inst,, Los Baños, Philippines.
Yoshida, S. 1981. Fundamentals of Rice Crop Science. IRRI, Los Baños, Philippines.

6
Hybrid Sterility in Rice—Its
Genetics and Implication to
Differentiation of Cultivated
Rice
H. Ikehashi*
INTRODUCTION
Chandratna (1964) reviewed the cytological basis of hybrid sterility. Oka
(1974) proposed a theory of gamete abortion by duplicate recessive
gamete lethal genes. Further studies by the present author and
coworkers have shown that an allelic interaction at a single locus is
responsible for gamete abortion.
In typical Indica-Japonica crosses, hybrid sterility is caused by an
allelic interaction at locus S-5, where Indica and Japónica cultivars carry
S-5' and S-5^ alleles respectively, while some cultivars have a neutral
allele, S-5”. The S-5VS-5^ genotype produces semi-sterile panicles due to
partial abortion of female gametes carrying the allele of S-5^ (Fig. 6.1 A).
Such abortion does not occur in S-5”/S-5' and 5-5“/S-5^ genotypes
(Big. 6.1B). The donor of S-5" is referred to as the widely compatible
variety (WCV) (ikehashi and Araki, 1986). As soon as the simple
monogenic nature of hybrid sterility was understood, it was applied to
Faculty of Agriculture, Kyoto Ur\iversity, Japan.

lio Rice Breeding and Genetics; Research Priorities and Challenges
hybrid rice breeding to enhance the level of heterosis.. This allele has
been incorporated into Indica and Japónica varieties to overcome hybrid
sterility and to realize pronounced heterosis in inter-subspecific hybrids
(Ikehashi, 1991; Yuana^ 1992; Zou et ah, 1992).
GENETICS OF HYBRID STERILITY
The allele S-5” has been effective for a large number of Indica and
Japónica crosses. Of the more than 10,000 varieties from China, only a few
show hybrid sterility in their crosses to WCVs (Wan and Ikehashi, 1995a).
However, in a broad range of varietal testing, such WCVs exhibited
hybrid sterility when crossed to varieties from the Indian subcontinent or
native varieties in China. Further genetic analyses of hybrid sterility gene
loci (HSGLi) were conducted subsequently.
A large number of three-way crosses (A/B//C) were made after
confirming that a hybrid A /C produced semisterile panicles and another
hybrid B /C was fertile. The progeny of A/B//C segregated into
semisterile plants as expected from A /C and fertile ones as expected
from B /C in a ratio of 1 ; 1. When the backcross A/C//C was made, the
progeny resulted in semisterile plants as expected from A /C and fertile
ones from C /C in a certain ratio. Thereafter, such genetic markers
CO segregating with semisterility were surveyed to identify a locus for
semisterility. In the backcrosses, plants were used as female to find
the distortion of marker genotypes which was caused by the abortion of
female gametes carrying one of the alleles. Table 6.1 shows an instance of
such an analysis in which the hybrid sterility in a cross between Jaya and
Ketan Nangkais cosegregated with two markers, Est-9 on chromosome 7
and Est-1 on chromosome 4, but not with Amp-3, which is the marker for
S-5 on chromosome 6. Allelic differences at the new locus were estimated
following the model of allelic interaction at S-5. For three given varieties.
Table 6.1
Distribution of spikelet fertility classified by marker genotype in
Jaya/Ketan Nangka/ZJCetan Nangka
Genotype Number of Plants in % Spikelet Fertility class Total Mean
S . 1
1 0 2 0 30 40 50 60 70 80 90 1 0 0
Est-9VEst-9^ 0 0 0 1 â 3 7 9 23 1 2 59»* 77.68*'"
Est-9^/Est-9^ 0 1 6 2 2 18 13 É 7 a 2 83 50.1
Est-lVEst-l° 0 0 0 1 1 1 9 12 2 2 11 57** 78.9**
Est-lV Est-l“ 0 1 6 23 2 0 15 4 4 2 a 85 49.8
Amp-3 VAmp-3^ 0 0 2 13 10 9 5 9 16 8 77 61.3
Amp-3^/Amp-3^ 0 1 4 11 11 7 8 7 15 6 75 60.4
** Shows significant difference between two genotypes at 1%.
Underlined numbers are assumed recombinants.

H. Ikehashi 111
©
0
0
©
VO
bi)
©
'a (1) tu
4-1 4-1
U cu
o E
,Q fd
(d tn
nH0)
(d44
E OJ
n E
o d
C Ö1
■n
rH tn
\ txT
p4 ‘H
4-)
tí u
M 4J
o tí
XI -H
X
cd
u
1
XI
« í

112 Rice Breeding and Genetics: Research Priorities and Challenges
A, B and N/ if a hybrid A /B showed gamete abortion at a HSGL S-X, but
N /A and N /B showed no distorted segregation of markers for S-X, the
variety N was determined to possess a neutral allele S-Xn at the new
locus.
IDENTIFICATION OF HYBRID STERIUTY GENE LOCI (HSGLi)
AND MARKERS
A number of new HSGLi were identified, as shown in Table 6.2. A locus
S-7 was detected in hybrids between aus varieties (summer rice in Indian
subcontinent) and some javanicas (Yanagihara et al., 1992). A lociis S-8
was detected in a hybrid between a Korean indica variety and some
javanicas (Wan et al, 1993). A locus S-9 was detected in hybrids between
aus varieties and some javanicas (Wan et al., 1996a). A locus S-15 was
found in hybrids between an aus variety Dular (WCV) and some IRRI
lines IR2061 (Wan et al,, 1996a). A locus S-16 was identified near Est-1 on
chromosome 1 in hybrids between Ketan Nangka and local varieties in
the Tai-hu Lake region of Yxmnan province in China (Wan and Ikehashi,
1995a). Locus S-17 (t) was identified in crosses between Penuh Baru II
and some japónicas (Wan and Ikehashi, 1995c). Isozyme analyses were
conducted according to Ishikawa et al. (1989) and Glaszmann et al.
(1988). From these genetic analyses, a number of tester varieties are being
developed. Such tester varieties maybe used to identify respective alleles
in other varieties.
Table 6,2
Loci for Hybrid Sterility
' Locus Chromosome Marker genes in order Crosses
1; S-5 6 wx, C, S-5, Amp-3, Est-2 Indica/Japónica
Pgi-2, RG213, alk
liv S-7 7 Rc, S-7, Est-9, rfs Aus/Javanica
j j
Ga-11, Acp-4
■F S-8 6 Cat-1, Pox-5, S-8 IR2061 /Javanica
s-9 , 4 Ph, Ig, Mal-1, Est-1, S-9 Aus/Javanica
S-15 1 2 Acp-1, Pox-2, S-15, Sdh-1 IR2061 /Dular(Aus)
'■':;j S-16 1 Linked Est-5 (15-20%) China N./Javanica
S-17(t) 12 Pox-2, S-15, Sdh-1, S-17(t) P.B. Il/Japonica
Initially, three alleles were identified at locus S-5. In the course of
further analysis, however, allelic differentiation at HSGLi was shown to
form a number of alleles at a single locus. Especially in Indian varieties,
more than five alleles were identified at S-7 using a set of testers. In the
survey of diversity of alleles at HSGLi and isozymes in Chinese varieties
and summer rice of India (Aus), the Indian cultivars showed the highest

H. Ikehashi 113
diversity in terms of alleles at HSGLi (Wan- and Ikehashi, 1997). For
instance, Basmati 370 showed several loci for hybrid sterility (Table 6.3),
viz.. S-8, S"7 and S-5 (Wan and Ikehashi, 1995b). In contrast, hybrid
sterility in typical indica-japonica crosses was predominantly controlled
by alleles at S-5.
HYBRID STERILITY IN POLLEN AND HYBRID RICE
Hybrid sterility is also expressed in pollen. Genetic analyses for hybrid
sterility in pollen were difficult, as there are a large number of pollen
genotypes in a single spikelet. Recently, it has become possible to
identify some loci for pollen sterility at a locus on chromosome 7 and
another on chromosome 12 (Wan and Ikehashi, 1996c). At ga-11 locus on
chromosome 7, pollen carrying an indica-type allele was found to abort
in the heterozygote. Thus, its mechanism is the same as that of female
gametes. It is therefore clear that hybrid sterility for pollen (male
gamete) and spikelet (female gamete) is independently controlled. This
is because a new gene responsible for both female and male abortion
may not be conserved.
Although a high level of heterosis was realized in indica-japonica
hybrids by utilizing WCG, their pollen fertility was low due to hybrid
sterility expressed in their pollen. The yield of hybrids was unstable in
adverse conditions due to poor pollen viability. Thus the importance of
WCG for pollen is recognized. The possible solution to this is to develop
indica-javanica hybrids, as many javanica varieties show normal pollen
fertility in their crosses to indicas and to japónicas ( Ikehashi and Araki,
1987). In China, a number of new rice hybrids using javanica and indica
varieties have been developed by the two-line system (Bai Delan and
Luo Xiaohe, 1996).
SYSTEMATIC ENHANCEMENT OF HETEROSIS
The hybrid sterility genes are diverse in Indian rice varieties (Table 6.3).
A range of potential restorers may be narrowed by hybrid sterility in
their crosses to cms lines, if the cms line contains non-neutral hybrid
sterility genes. It may therefore be advisable to incorporate maintainers
having Wide Compatibity gene (WCG) to use a wide range of restorers.
There should also be systematic enhancement of heterosis levels through
testing the combining ability of potential maintainers and restorers. The
significance of HSGLi in such a system remains to be investigated.

114 Rice Breeding and Genetics: Research Priorities and Challenges
Table 6.3
Distribution of spikelet fertility (SF) classified by marker
genotype in Basmati crosses
Genotype
Number of plants in % spikelet fertility
To 20 30 40 50 60 70 80 90 100
Total
Mean
Basmati 370 /IR36//IR36
Cat-lVCat-1^ 2 3 3 5 4 6 3 ^ 2 0 52.3**
Cat-lVCat-1^ 2 1 1. 1 1 2 1 17 18 16 61 75.8
SF was not differentiated at Est-2, Est-9, Est-1 and Sdh-1
Basmati 370/IR2061-628//Basmati 370
Cat-lVCat-l^ 1 1 1 1 0 2 8 10 8 6 38*»
Cat-1 VCat-1^ 3 2 2 4 7 9 a 8 g 4 55
SF was not differentiated at Est-2,Est-9, Est-1 and, Sdh-1
Basmati 370/IR2061-418//Basmati 370
76.3*»
51.7
Est-9VEst-9^ 2 2 3 3 6 3 14 a ■ S 4 50** 56.3**
Est-9 VEst-9^ 1 1 1 2 1 1 4 10 10 6 37 74.3
SF was not differentiated at Est-2,.Cat-l, Est-1 and Sdh-1
, Basmati 370/ Ketan Nangka / / Akihikari
Amp-3VAmp-3^ 0 1 0 0 3 2 14 11 7 5
Amp-3 VAmp-3^ 2 1 2 3 4 2 4 Z S 3
42
33
76.5**
52.3
Underlined numbers are assumed recombinants.
** Shows significant difference between two genotypes at 1%. (Wan and Ikehashi, 1995c)
DIFFERENTIATION OF HYBRID STERILITY GENES
An irradiated mutant Miyukimochi (Toda, 1982) was analyzed together
with its original variety, Toyonishiki. The semisterility in hybrids
between Toyonishiki and IR36 was caused by an allelic interaction of
S-5VS-5^, whereas the semisterility in hybrids between Miyukimochi
and IR36 was attributed to allelic interaction of both S-5VS-5^ and S-7VS-
7. Thus, the neutral S-7" in Toyonishiki was found to have mutated into
S-7'' by irradiation with Co^ (Wan and Ikehashi, 1996b).
Another case of mutational change of hybrid sterility allele was
found in an experimental line, 02428 from China, which possesses the
S-5” allele (Zou et al., 1992). The parents for 02428, Pangxiegu and
Jibangdao, were found to possess S-5^. The allele S-5” in 02428 was
considered to be induced from S-5^ by irradiation of the parents with
Co® (Wan and Ikehashi, 1996b).
The above incidences of mutation explain the way by which a new
allele is fixed. In the first case, a mutant allele S-7^ was induced in the
background of a neutral allele S-7” in Toyonishiki, and the heterozygote
produced S-7"/S-7" and S-7VS-7^ without showing sterility. In similar
ways, a number of new alleles might have been conserved in rice, as selffertilization
is predominant in this crop (Fig. 6.2). It is of interest that a
series of new alleles may easily be fixed under a genetic background of

H. Ikehashi 115
Q s - x v s - r
genetic drift
S-XVS-X"
< o
Fig. 6.2 Origin of hybrid sterility alleles at a locus.
neutral alleles^ while a different series of alleles under a different genetic
background. Thus, it is understandable that a varietal group contains a
series of alleles without revealing a high level of hybrid sterility as found
in intergroup hybrids. In test crosses among aus cultivars from the Indian
subcontinent, almost all the crosses showed no hybrid sterility, but the
same cultivars showed a wide range of hybrid sterility when testcrossed
with javanica, indicating their differentiation of alleles at HSGLs
as in the case of Jaya/Ketan Nangka (Table 6.2). Likewise, relatively few
cultivars showed hybrid sterility in crosses among javanica cultivars
(Ikehashi and Araki, 1987).
VARIETAL DIFFERENTIATION AND HYBRID STERILITY
Glaszmann (1987) has shown that there are predominantly indica and
japónica types in East Asia while there are a series of intermediate types
in the Indian subcontinent (Fig. 6.3), He suggested that they are
alternative gene pools. In the light of our analysis of hybrid sterility
genes, the hybrid sterility in Chinese cultivars was found to be caused
mostly by an allelic interaction at S-5, while the hybrid sterility genes in
Indian cultivars were found to be very much differentiated. The two
predominant types in East Asia can be understood if the two were
introduced originally from diverse sources as was the one in the Indian
subcontinent and propagated in the course of cultivation (Fig. 6.4).

116 Rice Breeding and Genetics: Research Priorities and Challenges
,__ I
' S
'-T'
&
/ \
! / \
c Q \
VII
Fig. 6.3
Distribution of 1688 Asian rice varieties in 6 varietal groups based on
isozyme variation at 15 loci. Groups are designated I to VI; class 0
corresponds to unclassified varieties from J.C. Glaszmann, 1987.
The mutation and fixation of hybrid sterility genes may not initially be
evident^ as indicated in Fig. 6.3. After isolation of some groups by human
activities and sporadic hybridization among them, such hybrid sterility
became apparent. Hybrid sterility in rice is cited as a major reproductive
barrier, which may cause isolation of varietal groups and lead to
differentiation. The hybrid sterility may be indicative of varietal
differentiation, which might be caused by the dissemination of a few
genetically narrow types in the expansion of rice cultivation. It is
important for breeders to understand that hybrid sterility is handled by
genetic manipulations. The concept of hybrid barrier should be critically
examined in the light of active breeding.
I. I

118 Rice Breeding and Genetics: Research Priorities and Challenges
References
\
i' I
Bai Delan and Luo Xiaohe, 1996. Liangyou Peite, a new tw6 -line hybrid rice released in
China. Rice Res. Notes. 21(1); 42-43.
Chandratna^ M.F. 1964. Genetics and Breeding of Rice. Chap.5, Longman, London, pp. 389.
Glaszmann, J.C. 1987. Theor, Appl. Genet., 74:21-30.
Glaszmann, J.C., Reyes, B.G. and Khush, G.S. 1988. Electrophoretic variation of isozymes in
plumules of rice (Oryza sativa L.)—a key to the identification of 76 alleles at 24 loci. IRRI
Research Paper Series No.134,
Ikehashi, H. 1991. Genetics of hybrid sterility in wide hybridization in rice. In: Biotechnology
in Agriculture and Forestry, Vol.l4. Rice, Y.P.S. Bajaj (ed.). Springer-Verlag, Berlin
Heidlberg, pp. 113-127.
Ikehashi, H. and Araki, H. 1986. Genetics of sterility in remote crosses of rice. In; Rice
Genetics, IRRI, Manila, Philippines, pp. 119-130.
Ikehashi, H. and Araki, H. 1987. Screening and genetic analysis of wide compatibility in
hybrids of distant crosses in rice Oryza sativa L. Tech. Bull. Trap. Agr. Res. Center, Japan,
No. 23.
Ichikawa, R., Mqrishima, H., Mori K. and Kinoshita, T. 1989. Chromosomal analysis of
isozyme loci and allelic expression at cellular level in rice. /. Fac. Agr. Hokkaido Univ.
64(1); 85-98.
Oka, H. 1974. Analysis of genes controlling Fj sterility in rice by the use of isogenic lines.
Genetics 77; 521-534.
Toda, M. 1982. The breeding of four new mutant varieties by gamma-rays in rice. Report of
symp. Breeding of Varieties by Use of Radiations. Gamma Field Symposia 21:7-15.
Wan, J., Yanagihara, S., Kato H. and Ikehashi, H. 1993. Multiple alleles at a new locus
causing hybrid sterility between Korean indica variety and a javanica variety in rice
{Oryza sativa L.). Jap. /. Breed. 43:507-^516.
Wan, J. and Ikehashi, H. 1995a. Identification of a new locus S-16 causing hybrid sterility in
native rice varieties {Oryza sativa L.) from Tai-hu lake region and Yunnan Province,"
China. Breed. Sci. 45:461-470.
Wan, J. and Ikehashi, H, 1995b. Loci for hybrid sterility in Basmati crosses. Inti. Rice Res.
Notes. 20:4.
Wan, J. and Ikehashi, H. 1995c. A new locus for hybrid sterility in remote crosses of
cultivated rice {Oryza sativa L.) Breed. Sci. 6 . Suppl. 2:191.
Wan, J., Yamaguchi, Y., Kato, H. and Ikehashi, H. 1996a. Two new loci for hybrid sterility in
rice {Oryza sativa L.). Theor. Appl. Genet. 92:183-190.
Wan, J. and Ikehashi, H. 1996b. Evidence for mutational origin of hybrid sterility genes in
rice (Oryza sativa L.). Breed. Sci. 46:169-174.
Wan, J. and Ikehashi, H, 1996c. A new locus for hybrid sterility in remote crosses of
cultivated rice {Oryza sativa L.)7. Breed. Sci. 46, Suppl. 1:87.
Wan, J. and Ikehashi, H. 1997. Identification of two types of differentiation in cultivated rice
{Oryza sativa L.) detected by polymorphism of isozymes and hybrid sterility. Euphytica,
94: 151-161. .
Yanagihara, S., Kato H. and Ikehashi, H. 1992. A new locus for multiple alleles causing
hybrid sterility between an Aus variety and Javanica varieties in rice (Oiyzfl sativa L.).
Jap. J. Breed. 42: 793-801.
Yuan, L. P. 1992. The strategy of the development of hybrid rice breeding. In: Current Status
of Two Line Hybrid Rice Research. L.P.Yuan, (ed.). Agrie. PubL, Ltd., Beijing, pp. 1-5.
Zou, J., Nie, Y., Pan, Q. and Fu, C. 1992. The tentative utilization of wide compatibility strain
02428 in Indica/Japónica rice. In: Current Status of Two Line Hybrid Rice Research,
L.P.Yuan, (ed.). Agrie. PubL, Ltd., Beijing, pp. 333-339.

A Critical Evaluation of
Current Breeding Strategies
M.J. Lawrence^ and D. Senadhira^
INTRODUCTION
It has been predicted that world production of rough rice will need to
rise by 70% in the next 35 years in order to keep up with the anticipated
growth in the human population and income-induced demand for food
(IRRI, 1993). There has been no increase in the area planted with rice
since 1980, and increasing urbanization and industrialization can be
expected to reduce this area (FAO, 1988). It follows that only raising the
average yields of crops grown in existing areas can increase rice
production. Since 73% of current rice production comes from crops
raised under irrigation, most of the required increase will have to come
from new, high-yielding and stable varieties that have been bred for this
ecosystem.
In principle, an increase of 70% should be achievable, not only
because both the world and Asian production of rice has, as a result of
the work of agronomists and breeders, doubled over the past 30 years,
with only a modest concomitant increase in the area devoted to the
irrigated crop but, more particularly, because national yields, in the
^Plant Genetics Group, School of Biological Sciences^ University of Birmingham, Birmingham
B15 2TT, UK
^ Plant Breeding, Genetics and Biochemistry Division, International Rice Research Institute,
PO Box 933,1099 Manila, Philippines

120 Rice Breeding and Genetics: Research Priorities and Challenges
I■ 'i
I !■
majority of countries, are still well below the maximum possible with
existing cultivars. In practice, however, it may be more difficult to
increase rice production by this amount. First, whereas most modern
indica cultivars are bred for use in transplanted crops, anticipated rising
costs of labor will make this traditional practice increasingly uneconomic.
New varieties, therefore, will have to be bred for use in directly
sown crops; this will require a modification of the present IR8 ideotype
towards plants which produce fewer tillers, all of which are fertile.
Agronomists may also have to devise better ways of managing fertilizers
and controlling weeds in the crop. Secondly, and more importantly
there appears to be a widespread belief among breeders that modern
indica cultivars that have been bred for use as transplanted crops in the
lowland tropics have reached a yield potential barrier of 10 t ha"^. The
evidence on which this belief is based is that while recently released
cultivars of this type, such as IR64 and IR72, achieve higher yields in
agricultural environments than the pioneering variety of this ideotype,
IR8, released in 1966, the yield of the latter differs little from that of the
former in experimental environments in which crops are protected from
attack by pests and diseases (Flinn et ah, 1982; Cassman et al., 1994;
Kropff et al, 1994). Hence the superior performance of more recent
varieties in agricultural environments is chiefly due to their genetic
resistance to a number of important pests and diseases, which IR8 lacks;
that is, their superiority in this respect lies in the fact that they are able to
avoid losses of yield, and thus have a greater yield stability over seasons
and locations. Nevertheless, these observations suggest that despite 30
years of attempting to breed for increased yield, the yield potential of
the rice crop has remained more or less static.
In an attempt to overcome this serious problem, scientists at the
International Rice Research Institute (IRRI) have initiated two
alternative breeding programs. One involves a radical redefinition of
ideotype, known as new plant type, in which recombinant inbred lines
are extracted from crosses between temperate and tropical japónica
cultivars, and the other, the production of hybrid varieties, initially from
crosses between indica parents, but later from indica x japónica crosses
(IRRI, 1989; Khush et al, 1994; Peng et al, 1994). There may be a number
of other reasons for preferring these crosses to those between indica
parents, among which is the desire to reduce panicle number per plant,
and to increase culm thickness and number of spikelets per panicle in a
directly seeded crop, as in the new plant type program, or to exploit
heterosis for yield in hybrid varieties. Nevertheless, the cause or causes
of the yield barrier with indica cultivars appears to have received
surprisingly little attention. So we shall turn first to a discussion of
possible explanations of this problem before going on to review the

M.J. Lawrence and D. Senadhira 121
results of some experiments which, though undertaken with this
particular problem in mind, have implications for the improvement of
the rice crop in general.
THREE POSSIBLE EXPLANATIONS OF THE STASIS FOR YIELD
There are three possible explanations for this lack of progress with yield.
First, modern indica cultivars of type IRS may have a narrow genetic
base for yield and other characters of interest. If so, further substantial
responses to selection for these characters are unlikely because most of
the increasing alleles at the loci controlling them have become fixed as a
result of past selection or, perhaps, because of the effects of genetic drift
causing random fixation of alleles in populations of small effective size
(an effect usually overlooked in the literature). In these circumstances,
selection has reached a limit, which, in the absence of mutation or
introgression of genes from other sources, cannot be breached (Falconer
and Mackay, 1996). Were this the case, no further explanation Would be
necessary. The belief that this first explanation is true appears to have
been one of the reasons underlying the decision to initiate the new plant
type program at IRRI.
Second, though considerable genetic variation for these characters
is, in fact, still present, conventional breeding procedures have failed to
exploit it. These methods have, of course, been successfully employed to
produce the great majority of all varieties of small-grain cereals,
including rice. Their success rate per cross is low, however, so breeders
have to make many hrmdreds of crosses each season to ensure that some
promising lines ultimately emerge from their programs. These methods,
therefore, chief among which is the pedigree method, cannot, on this
evidence alone, be regarded as Very efficient. This possibility appears
not to have been considered by breeders at IRRI.
Third, while conventional breeding procedures may be reasonably
efficient in exploiting genetical variation when selecting for enhanced
performance with single characters, this is not the case when attempting
to select for.a number of characters of interest simultaneously. Thus, for
an aggregate phenotype consisting of k completely uncorrelated and
equally important components, if p is the overall proportion of individuals
to he selected, the proportion, p,, selected for each component
considered independently of every other (independently culling levels)
is equal to the /cth root of p„ Unless k is very small, it is not
possible to carry out more than rather weak selection on these
components; for example, if p is 0.05 or 5%, then for an aggregate
phenotype consisting of three components, the selection pressure on

122 Rice Breeding and Genetics: Research Priorities and Challenges
each, measured by pj, can be no greater than 0.37 or 37%. This problem,
for which unfortunately there is no general solution, is, of course,
encountered in virtually all breeding programs; if the second
explanation also holds, this problem will be worse. In the case of indica
cultivars, it is possible that in introgressing genes for disease and pest
resistance from very low yielding donors, via wide hybridization with
other Oryza species, the selection pressure on yield was relaxed in order
to ensure that some resistant recombinant inbred lines were obtained.
Investigating the First Possibility
If the first explanation is true, evidence for heritable variation for
quantitative characters of interest, including yield and its components,
should be scarce in all crosses between indica parents. In consequence,
the heritability of these characters should be low and it should be
difficult to extract a recombinant inbred line whose performance exceeds
that of the better parent from any of these crosses. If, on the other hand,
it turns out that this is not the case, one or other of the other explanations
(or both) must hold.
The trial designs used by breeders are not capable of yielding
estimates of the genetic parameters that are required to test the first
possibility because the material raised in them is neither randomized
nor appropriately replicated. Perera et al. (1997) investigated the
potential of a total of ten crosses between indica parents to yield
transgressive segregants with respect to twelve quantitative characters
in three experiments that were designed with this purpose in mind. All
of these parents, which between them included all three of the age
classes of the crop, were of improved plant type with a relatively high
yield potential. Seven of the crosses (crosses 1-7) were effectively taken
at random from conventional breeding programs at the Rice Research
and Development Institute at Batalagoda in Sri Lanka where these
experiments were carried out; the remaining three (crosses 8 - 1 0 ) were
made with the specific purpose of maximizing response to selection for
yield and other characters of interest. In each of these experiments, the
families of the basic (P^, P 2 , Fj, F2 , Bj and B2 ) and F3 generations of each
cross were raised in completely randomized blocks and plants scored
individually for each character.
The results obtained from these experiments leave little doubt of the
presence of considerable amount of additive genetical variation in each
cross for each of the characters scored for the following reasons. First,
the parents differed significantly for 63 out of a total of 105 combinations
of crosses and characters; in addition, the FiS displayed significant
better parent heterosis for 23 Of these combinations. Overall, 78 of these
combinations (74%) indicated that the parents of these crosses differed

M,J. Lawrence and D. Senadhira 123
genetically; this is the simplest and most direct evidence of ttie presence
of heritable variation in these crosses.
Second, variance analysis of the F3 families of these crosses showed
that with three exceptions only, all characters were heritable. However,
the mean scores of the parents differed significantly for two of these
exceptional characters and the Fj displayed significant better parent
heterosis for the third. Taken as a whole, therefore, these results leave
little doubt that all characters in all crosses were heritable. Estimates of
their heritability (Table 7.1) showed that nearly one-third were 50% or
greater and that less than one-fifth were below 2 0 %.
Table 7,1
Estimâtes of the heritability (7o) of characters obtained from the
variance components of variance analysis of F3 families
Cross TN DH I n* CL PL NPP PW GW' NGP GY NEP NSP
1 25 46 ■23 48 55 48 45 23 39 - -
2 6 8 63 48 36 33 24 54 31 1 0 0 - -
3 28 50 1 0 0 (5) 90 54 63 30 1 0 0 _ - -
4 17 6 8 30 25 31 2 0 31 16 76 - - -
5 64 6 8 39 56 64 75 81 43 1 0 0 “ - -
6 15 63 50 37 21 38 47 92 32 16 63 42
7 (2 ) 82 44 34 19 5 17 61 19 11 36 32
8 11 93 70 40 9 (2 ) 15 30 1 2 31 28 1 2
9 67 46 29 46 21 59 29 58 42 44 29 35
10 11 63 67 96 31 21 28 60 6 43 24 9
Note: LOb2029xBg34-6(3-^ months);2,Bg380-2xBg34-6(3-L months);82-662 x 82-618(4
4 months); 4, Bg380-2 x 82^662 {4-4-j- months); 5,82-1799xIR50 (3 months); 6, Bg850
X1R50 (3 y months); 88-5328 x Ob2552 (4-4^ months); 8,Bg34-8 x IR58 (3 months); 9,
Bg94-1 X IR62 (3-1-months); and 10, Bg90-2 x IR72 (4 -4 ^ months). Key to the
characters scored TN = tiller number, DH = days to heading, HT = height, CL = culm
length, PL = panicle length, NPP = number of panicles per plant, PW = panicle weight,
GW = thousand grain weight, NGP = number of grains per panicle, NEP = number of
empty spikelets per panicle and NSP = number of spikelets per panicle. The F3 families
of Crosses 1-5 were raised in one trial, those of 6 and 7 in a second and those of Crosses
8-10 in a third. The entries in parentheses indicate characters for which although there
was no evidence of heritable variation from variance analysis of P3 families, were
clearly heritable on other evidence (see text). Parents of crosses with age class in parentheses:
l,Ob2029X
Third, predictions of the proportion of recombinant inbred lines that
can be extracted by single seed descent from these crosses, whose means
are equal to or greater than that of the better parent (or that of the Fj
where this shows significant better parent heterosis), showed that it
should be possible to obtain one or more such lines for most characters if
100 were produced from each cross. To be 99% certain pf obtaining one f:

124 Rice Breeding and Genetics: Research Priorities and Challenges
or more lines that achieved the desired targets for a minority of
characters, however, these predictions indicated that it might be
necessary to raise as many as 500 such lines by single seed descent. Only
four of these predictions suggested that there was no prospect of
achieving the desired targets for the characters concerned. These crosses
were, therefore, clearly capable of yielding superior transgressive
segregants for the great majority of characters.
Fourth, the proportions of F3 families of crosses 8 , 9, and 10 whose
means achieved the desired targets for characters of interest were
broadly consistent (though, as expected, on average a little lower) with
the corresponding proportions of recombinant inbred lines expected to
achieve these targets (Table 7.2). If the means of F3 families display
useful transgressive variation, those of their F descendants are virtually
certain to do so.
Table 7 .2
The predicted proportion (P) of recombinant inbred lines that can be
extracted from Crosses 8-10 by single-seed descent whose means are
expected to equal or exceed the indicated targets and the proportion
of F3 families {Pf} whose means achieved these targets.
K Cross Statistic PL NPF PW GW NGP GY NEP NSP
i 8 Target ¿26.0 ^ 35.8 >3.8 >26.7 > 130.6 >94.7 35,8 >4.0 >26.5 ¿ 132.8 ¿ 93.8 S30.6 ¿ 167.5
1 , ; P 12% 25% 5% 11% 34% 31% 26% 42%
1 i' 15% 15% 8% 5% 33% 28% 23% 35%
i . 10 Target >29.2 >33.2 >5.1 ¿29.4 ¿ 174.9 ¿ 136.0 ^21.9 > 2 1 1 .2
■i !■ M P 30% 31% 4% 10% 5% 18% 2% 15%
y
2
^ _____
1% 32% 5% 5% 1 1% 8% 0% 13%
Note: Targets were taken as the means of the higher scoring parent or that of the Fj, when this
was greater, except for NEP where the target was the mean of the lower scoring parent or
that of the Fj when this was lower.
Fifth, these crosses had considerable potential for yield. The
predicted yields in tons per hectare of the parents and F^s of crosses 8 ,9 ,
and 1 0 , together with those of their highest scoring F3 family and the
predicted yield of the best 1 % of the recombinant inbred lines that can be
extracted by single seed descent from these crosses, are shown in Table
7.3. The entries in this Table leave little doubt that it should be possible
to extract recombinant inbred lines from each of these crosses whose
performance is appreciably better than that of their F^. The predictions
for Cross 10 are of special interest because it is of the same age class (4-
4 y months) as the crosses in IRRI's new plant type program. We note
that the predicted yield of the best F3 family is, at 9.65 t ha'^ only just
short of the assumed yield barrier of 1 0 t ha"^ and, in particular, that of

M J. Lawrence and D. Senadhira 125
the top 1% of recombinant inbred lines is at 11.37 t ha'^ comfortable
greater than this barrier.
Table 7.3
Predicted yields (t ha‘^) of the parents of Crosses 8-10 (Bg and IR),
their Fj, their highest scoring F3 family and those of the top 1% of
the recombinant inbred lines (RlLs) that could be extracted from
them.
Family Cross 8 Cross 9 Cross 10
Bg 5.16 5.30 8.31
IR 4.76 4.94 6.89
Fi 5.92 5.86 8.50
Best F3 6.27 6.79 9.65
Topl%ofRILs 6.61 8.10 11.37
The results of this investigation are inconsistent with the notion that
there is a shortage of exploitable genetical variation for yield and other
quantitative characters among modern indica cultivars and, in
particular, are not consistent with the belief that the crop has reached a
yield potential barrier of 10 t ha'^. The possibility that the crop has a
narrow genetic base can, therefore, be discounted; it follows that the
inability of breeders to breach this barrier must have some other cause.
Investigating the Second Possibility
Conventional breeding methods involve selection in the early
generations of crosses. In the case of the widely used pedigree method,
selection is carried out in every generation, whereas with the bulk
method, this is usually not initiated before the F5 generation. The
modified pedigree method is intermediate in this respect, weak negative
selection (perhaps better described as culling) being carried out on bulk
populations of the P2 F3 generations of crosses, before selection of a
similar type to that employed with the pedigree method is practiced in
the F4 and Fg generations. The selection practiced for characters such as
height and days to maturity is of the stabilizing type, with the aim of
choosing individuals that have a similar phenotype to that of
recommended varieties that are included in trials for this purpose.
Selection for plants with large panicles bearing a large number of
spikelets, on the other hand, is directional. All of this selection is visual
and therefore indirect; yield, for example, is not usually measured
directly much before the F^ generations of pedigrees when individuals
are more or less homozygous.
Now, if this early generation selection is effective, the recombinant
inbred lines produced by these methods are expected to be better than
those produced by single-seed descent (SSD), in which no selection is
practiced during the course of inbreeding. The SSD method, therefore,

126 Rice Breeding and Genetics: Research Priorities and Challenges
can be used as a control against which the efficiency of conventional
methods can be assessed. Fahim et al. (1997) compared the pedigree,
modified pedigree and bulk methods with SSD in terms of their relative
efficiency in exploiting heritable variation for yield and other
quantitative characters in Crosses 6 and 7. The number of Fg families
produced by each of these methods in each cross is shown in Table 7.4.
The means of the families produced by the conventional methods were,
on average, greater in both crosses than those produced by SSD for
many of the characters, including grain yield, for which greater
expression was desired. However, though without exception the best
families produced by each of these methods achieved the desired target
for every character considered on its own, none were significantly better
than the best family produced by SSD (Table 7.5). The chief effect of the
selection carried out with the conventional methods was, therefore, to
cull plants of poor performance (negative selection) rather than to select
those of superior performance (positive selection).
Table 7.4 The number of
families produced by the four breeding methods in each cross
Method Cross 6 Cross 7
Pedigree (P)
Modified pedigree (MP)
6 uIk(B)
Single-seed descent (SSD)
51
14
25
100
22
12
30
100
Total 190 164
Table 7.5
Mean Scores of the best Fg families produced by each method for
each character for which improvement was desired from each cross
Cross PL NPP GY GW PW NGP NEP NSP
Target ä25.9 a 27.2 >87.2 >23.8 a 4.45 a 175.8

M.J. Lawrence and D. Senadhira 127
each cross. If this comparison was confined to. the costs incurred from
the p2 to the Fs of these pedigrees, SSD would be much less costly than
the other methods because with SSD, generations were advanced in the
nursery, rather than in the field. The audit of these costs also did not
include the salary of the plant breeder who carried out selection in the
conventional methods. Again, for convenience, the pedigrees of these
crosses were advanced synchronously at the rate of two generations per
annum with all four methods; in practice, it should be possible to raise
three or more generations per annum with SSD, giving it a further
advantage over conventional methods. Lastly, it can be argued that this
comparison is biased against SSD because whereas several thousand
plants were raised in each generation with the conventional methods,
only 100 lineages were advanced to the Fg generation with SSD; if, say,
500 such lines were raised in the trial, it is likely that the best would
have significantly outperformed the best lines produced by the other
methods.
Despite these difficulties, the results of this comparison leave little
doubt that SSD was at least as effective as conventional methods in
exploiting the heritable variation in these crosses. Though, as far as we
are aware, this is the first such investigation with rice, a number of other
investigation into the efficiency of early generation selection have been
carried out. Among these are those of Empig and Fehr (1971), and
Boerma and Cooper (1975) with soybean; Casali and Tigchelaar (1975 a)
with tomato; Kaufmann (1971) with oats; Park et al. (1976) with barley;
and Knott and Kumar (1975), Tee and Qualset (1975) and Oeveren (1992)
with wheat. Casali and Tigchelaar (1975b), Cornish (1990a, b) and
Oeveren and Stam (1992) have also investigated this question by
computer simulation. Though the methods and designs used in these
previous investigations differ in some important details from those
employed by Fahim et al. (1997), there is general agreement that, in
terms of the production of the best lines, SSD is, at worst, only slightly
less efficient than other methods, is frequently superior, and is
potentially more rapid and cost effective than the latter.
The conclusion that emerges from this investigation of the second
possibility is obvious, namely, that the inability of breeders to breach the
1 0 t ha"^ yield barrier with indica cultivars is almost certainly due in
part, at least, to the use of inefficient breeding procedures. If, therefore,
the attempt to practice selection in the early generations of pedigrees
were to be abandoned, a larger number of recombinant inbred lines
could be produced by SSD, for a given expenditure of effort, than with
present methods, wtuch would increase the probability of extracting
superior transgressive segregants from crosses.

123 Rice Breeding and Genetics; Research Priorities and Challenges
Investigating the Third Possibility
It is not possible^ as pointed out earlier, to carry out more than rather
weak selection on the components of an aggregate phenotype unless the
number of such comppnents is small (or the population of individuals
that are candidates for selection is very large), if the proportion of
individuals ultimately selected is to be held at a reasonable level. In the
simple, three-component example used to illustrate this point, it was
assumed that the components were uncorrelated; this is, of course, very
unlikely in practice. Suppose that the desired gain for each of these
components is in the increasing direction. Then, if these components are
positively correlated, the response to selection on the aggregate
phenotype will be better than what our calculation indicated. If on the
other hand, one of the components is negatively correlated with the
others, the response expected will be less than if they were imcorrelated.
The sign and .magnitude of these correlations, therefore, determines the
overall response to selection on the aggregate phenotype.
Though selection has to be made with respect to the phenotypes of
individuals, it is the genetic correlation between components that
influences the. magnitude of this response. There are two causes of
genetic correlation—pleiotropy and linkage disequilibrium (Falconer
and Mackay, 1996). When the same set of genes determines two
characters, they are likely to be pleiotropically correlated; we expect
tiller and panicle number, for example, to be pleiotropically related.
Both Falconer and Mackay (1996) and Simmonds (1979) state that
pleiotropy is the chief cause of genetic correlation, because in
populations of individuals that mate at random, linkage disequilibrium,
especially if the loci concerned are not linked in their inheritance, is
expected to decay quite quickly over generations. There are good
reasons, however, for expecting linkage disequilibrium to be an
important cause of genetic correlation between characters in the crosses
handled by breeders of self-pollinating crop plant species. Thus, any
cross between a pair of homozygous parents is bound to generate
linkage disequilibrium with respect to loci for which they differ.
Suppose that the genotype of one parent with respect to a pair of loci is
A"'' and that of the other parent is A'*^A^B’*'B‘*'. Then, initially, the
only gametes present are (those of the first parent) and
(those of the second parent). In the next generation, however, the
missing pair of gametes, A'’B‘^ and A"^B^ are generated as a result of
recombination and segregation at meiosis in the Fj double heterozygote,
A'^'B^/A^B’*'. If the loci are not linked, the frequencies of the four-gamete
types (haplotypes) are expected to be equal and linkage equilibrium will
have been restored in a single generation. If, on the other hand, the loci

M J. Lawrence and D. Senadhira 129
are linked in their inheritance, the frequencies of the parental gamete
types will be higher than those of the recombinant types by an amount
depending on how closely they are linked. Furthermore, this linkage
disequilibrium of genes that are linked in their inheritance will decay
much more slowly over the generations of the selfing series than under
random mating because, since the frequency of heterozygotes halves
with each generation of the former, opportunities for recombination
become increasingly less common. We have gone into this important
matter in some detail because it appears to have been rarely considered
in the literature.
Now, if the genetic correlation between a pair of characters is,
allowing for some sampling variation, more or less constant in
magnitude and sign over crosses, it is likely that the chief cause of this
correlation is pleiotropy. But if, on the other hand, thé chief cause of
genetic correlation is the linkage disequilibrium of genes that are linked
in their inheritance, the magnitude and sign of the correlation would be
expected to vary over crosses, so that, in principle, breeders could select
those crosses in which the genetic correlations between characters of
interest were most favourable to their objectives. Sriyoheswaran (1995)
obtained estimates of the genetic correlations between the characters
scored in Crosses 8-10 from the components of the mean products of the
analysis of covariance of the data from the F3 families of these crosses.
Bearing in mind that the proportion of variation of one character caused
by its relationship with another is the square of the correlation between
them, attention may be confined to those pairs of characters for which
the absolute value of the estimate is equal to or greater than 0.5. Figure
7.1 shows the genetic correlations between characters in these crosses
which meet this criterion, the great majority of which were significantly
different from zero. The correlations between five pairs of characters
(TN and NPP; PW and NGP; NSP and H, PW and NEP) are consistent
over crosses in terms of their sign and relative magnitude; in addition, a
sixth pair (NEP and NSP) only just fails to fall into this category, the
correlation between these characters in Cross 8 being 0.48. The cause of
the genetic correlations between these characters (most of which are, of
course, expected) can, therefore, be attributed to pleiotropy. Apart from
these, the most striking feature of the correlations shown in Figure 7.1 is
the extent to which they vary over crosses. While it is not possible to
definitely njle out the possibility that the genetic correlations between
these characters is due to pleiotropy, because different genes will have
segregated in these crosses (see Falconer and Mackay, 1996 on this
point), the most reasonable interpretation of this variation over crosses
is that the chief cause of genetic correlation between these draracters is
the linkage disequilibrium of genes that are linked in their inheritance.

m
M J. Lawrence and D. Senadhira 131
It is worth pointing out that this variation in, the magnitude and, in
one case (GY and NSP), also of sign, is not a consequence of the decision
to consider only those correlations whose absolute magnitude is equal to
or greater than 0.5, since the great majority of estimates not included in
the Figure are well below this value and hence are not significantly
different from zero. For example, while the correlation between GY and
TN was 0.95 in Cross 9 and 0.70 in Cross 10, it was only 0,11 in Cross 8 .
Again, while GY was highly and positively correlated to H in Cross 10
(0,77), there was little evidence that these characters were correlated in
Cross 8 or 9 for which the estimates were only 0.05 and 0.22,
respectively. Indeed, for GY and PW, while the correlation between
these characters was 0.70 and 0.75 in Crosses 8 and 10 respectively, it
was - 0.45 in Cross 9, an estimate which was just significantly different
from zero; this is one of the very few cases where truncation may have
concealed a relationship of potential interest.
This variation between crosses with respect to the magnitude and
sign of the genetic correlations between characters has been routinely
observed in other rice crosses of the indica type (see Perera, 1985 for
Crosses 1-5), in a pair of new plant type crosses (Bentota et ahf 1997),
and also in five spring barley crosses (Thomas and Tapsell, 1985). The
ubiquity of this variation strongly supports the conclusion that the chief
cause of genetic correlation between characters in pedigrees founded by
crosses between pairs of inbred lines is the linkage disequilibrium
between the genes that determine these characters that are linked in
their inheritance. This, in turn, suggests that breeders might be able to
select crosses not ordy in terms of their capacity to yield superior
transgressive segregants with respect to single characters of interest, but
also those in which the genetic correlations between characters are
potentially favorable to the breeders objectives. Crosses which fulfill the
first of these requirements in which, in addition, characters for which
greater expression is desired are substantially and positively correlated,
and the correlations between this group of characters and those, such as
days to maturity, which breeders wish to maintain within an interval,
are zero or small, are of greater interest than crosses in which these
additional requirements are not met.
A BIOMETRICAL BREEDING PROCEDURE
Perera et al. (1997) and Fahim et al. (1997) carried out investigations for
the specific purpose of identifying the cause of the inability of breeders
to breach an assumed yield potential barrier of 1 0 t ha"* with indica
cultivars bred for use in the lowland tropics. They showed that there is

132 Rice Breeding and Genetics: Research Priorities and Challenges
no evidence of a shortage of exploitable genetical variation for yield and
other characters of interest in such cultivars and that the chief cause of
the failure to make progress with yield is almost certainly the use of
inefficient breeding procedures coupled with the well-known difficulty
of making progress with multitrait selection. These results; however,
have broader implications, because the procedures used indicate how
present breeding methods could be modified to make them more
efficient.
We have proposed elsewhere (Lawrence and Senadhira 1997) a
biometrical procedure which involves selection carried out at the Fj, F3 ,
and Fg generations of pedigrees founded on single crosses between
inbred parents, the first two stages involving selection between
pedigrees and the third, selection within pedigrees that survive the first
two stages of scrutiny. The purpose of the first stage is to identify
crosses whose Fi progeny display better-parent heterosis for yield and
other characters for which improvement is desired because, in the
absence of overdominance (for which it is now agreed there is very little
evidence; see Jinks, 1983), heterosis can occur only when the increasing
genes for which the parents differ are at least partially dispersed
between them. But such crosses are those most capable of displaying
transgressive variation in later generations; heterosis serves as a means
of recognizing these crosses. A trial which includes the parents and the
Fj^s of a number of crosses would allow breeders to identify those which
are most promising in this respect and to discard the rest.
The purpose of the P3 trial is to provide answers to four important
questions about crosses. First, are all characters of interest heritable?
Second, is the cross predicted to yield a reasonable proportion of
superior recombinant inbred lines for such characters?
Third, do the means of F3 families display useful transgressive
variation? Fourth, is the pattern of genetic correlations between
characters favorable to the objectives of the program in improving all
components of the desired aggregate phenotype? If the answer to any of
these questions is no, breeders would need to consider whether it was
worthwhile advancing the pedigree further.
Any cross which survives the first two stages of this objective and
systematic procedure is likely to produce superior recombinant inbred
lines for each of a number of characters considered singly, and also
simultaneously for some if not all, of all the characters for which
improvement is desired. The purpose of the third and final, F^ stage is to
identify such lines that are extracted from crosses by single-seed descent.
For reasons given earlier, there is little point in attempting to carry out
selection for quantitative characters during the course of inbreeding; it
is better to defer selection within pedigrees until such time that

M J. Lawrence and D. Senadhira 133
individuals are^.more or less homozygous and to redirect effort and
resources to maximizing the number of recombinant inbred lines
produced from each cross.
It must be emphasized that the trials at each of the three stages of
this new procedure are capable of providing objective and unbiased
information on the genetical variation segregating in crosses, if, and
only if, the entries in them are appropriately replicated and rar^domized,
and plants are scored by measuring them directly for characters of
interest. Since none of these crucial requirements are fulfilled at any of
the early stages in conventional breeding programs, the implementation
of this new breeding strategy would, of course, involve a considerable
departure from current practice. - .
THE RELIABILITY OF PREDICTIONS
Prediction of the proportion of recombinant inbred lines that can be
extracted from a cross by single-seed descent, whose means equal or
exceed any desired standard, involves estimation of the mean, m and
variance, D (or, in the alternative notation, 2VA*; see Kearsey and Pooni
1996) of the distribution of these lines from the early generations of the
pedigree (Jinks and Pooni, 1976). Calculation of the one-tailed normal
deviate, using these estimates then allows one to determine this
proportion from standard statistical tables. Estimates of these
parameters can be obtained from a variety of sources (Jinks and Pooni
1980). Perera et al. (1997) chose to derive these estimates from P3 families
on the grounds that families of this generation have to be raised in order
to advance pedigrees, whereas those of triple test crosses (Kearsey and
Jinks, 1968), for example, do not. How reliable are these predictions?
Table 7.6 shows the predicted and actual proportions of recombinant
inbred lines extracted by SSD from Crosses 6 and 7 whose
performance achieves the target for each of the characters for which
improvement was desired (Fahim et ah, 1997). It is evident that there is
reasonable agreement between the predicted and observed proportions
over characters. But the families in this case were raised in the same
trial as the F3 families from which estimates of m and D were obtained.
In practice, of course an Fg trial is likely to be held at least one year after
the corresponding F 3 trial; that is, these trials will be carried out in
different environments. Any difference between these environments
which has the effect of simply adding or subtracting a more or less
constant amount to the scores of individuals should be easily detected
from a. comparison of the grand means of the two trials (i.e. P3 and F 5 )
and accommodated, when the scores of the parents of the cross have

134 Rice Breeding and Genetics: Research Priorities and Challenges
been used to define the targets for improvement by including tihem in
the Fg trial. A similar argument holds for that type of genotypeenvironment
interaction which expands or contracts the observed
variance of the distribution of recombinant inbred lines about their
mean relative to the estimate of this variance obtained from F3 families,
because the rank order of genotypes, allowing for some sampling
variation, is not expected to change much in these circumstances.
Potentially the most serious form of genotype-environment interaction
is that which changes the rank order of genotypes over environments
(often referred to as crossover interaction). There are two reasons,
however, for believing that this type of interaction is unlikely to cause a
serious loss of reliability of the prediction of the proportion of superior
lines that can be extracted from a cross. First, while there is no doubt
that genotype-environment interaction of this type occurs in rice, as in
other crops, nearly all of the relevant evidence appears to come from
multilocational trials. Breeders, on the other hand, will almost certainly
carry out trials in the same location as the F3 trials; differences
between environments in the same location are expected to be smaller
than those between the environments of different locations, especially
for irrigated crops. Second, the genotypes most vulnerable to changes in
their rank order are those that determine intermediate phenot3^es;
extreme lines, those that fall into the tails of the distribution, are less
likely to suffer these changes. In consequence, neither the mean, nor the
variance of the distribution of recombinant inbred lines with respect to a
quantitative character is expected to be much affected by genotypeenvironment
interaction of this type. In general, therefore, genotypeenvironment
interaction is not expected to markedly affect the accuracy
of single character predictions of the proportion of superior inbred lines
that can be extracted from a cross; there is now a considerable body of
empirical evidence that supports this conclusion.
Table 7.6
Predicted and actual proportions (%) of the recombinant inbred lines
extracted from Crosses 6 and 7 whose performance achieved the
target for each of the characters for which improvement was desired.
Cross Proportion PL NPP GY GW PW NGP NEP NSP
6 Predicted 53% 61% 29% 27% 3% 3% 32% 8 %
Actual 38% 37% 25%= 45% 7% 6 % 31% 9%
7 Predicted 6 6 % 40% 16% 83% 35% 1 1 % 40% 23%
Actual 58% 2 2 % 9% 80% 35% 15% 47% 19%
Note: The Targets were the means of the better parent of each cross for the character
concerned.
This conclusion is not expected to hold, however, when a major gene
for, say, disease resistance is segregating in a cross. Clearly, when the

M.J, Lawrence and D. Senadhira 135
incidence of the disease in question is negligible in the F 3 trial but
appreciable in the trial, genotypes which determine high-yielding
phenotypes in the former are not expected to do so in the latter if they
are susceptible to this disease. If resistance is recessive, it would be
possible to screen the individuals of the F2 generation of the cross,
provided that both the expressivity and the penetrance of the gene is
complete, and the screening procedure used allows no escapes. If these
conditions can be met, then single-seed descent could be initiated from
the sub-set of F2 individuals that are resistant to the disease. Because
resistance is usually dominant, however, it would be necessary to carry
out a preliminary screen at F2 and a final screen at F^ on a subset of
seedlings immediately prior to the third stage trial, in order to identify
recombinant inbred lines that homozygous for resistance (see Lawrence
and Senadhira, 1998 for details). In all other circumstances, selection
within pedigrees should be delayed until the F^ generation.
SELECTING FOR STABILITY OF PERFORMANCE
The ideal variety is one that performs well in the full range of
environments in which it can reasonably be grown; that is, it displays
little genotype-environment interaction over these environments so its
performance is stable. There are two points worth making about stability
of performance. First, experiments with Nicotiana rusUca have shown
that it is possible to successfully select for all four combinations of the
extremes of mean performance and environmental sensitivity so as to
obtain recombinant inbred lines that are high-high, high-low, low-high
and low-low for these characters respectively (Brunmpton et al, 1977;
Jinks et al, 1977, Boughey and Jinks, 1978; Boughey et al, 1978; Jinks and
Pooni, 1987). This indicates that although these characters are often
positively correlated before selection, they must be, at least partially,
controlled by different sets of genes. There is, of course, no reason for
supposing that this is not true of other species also. This suggests that it
might be desirable to carry out Fg trials in a limited number of locations
that represent the range of environments in which the crop can be
reasonably grown, rather than a single location as with current practice.
If this could be done it would be possible to identify recombinant inbred
lines that have a high and stable yield much earlier than at present.
Gravois et al (1990), using Shukla's (1972) stability statistics and Kang's
(1988) rank sum method, have shown how high-yielding and stable
genotypes can be identified in rice.
Second, the environments used by breeders to evaluate crosses are
irearly always considerably better than those in which farmers raise
their crops; on these grounds alone, a case can be made for carrying out

136 ’ Rice Breeding and Genetics; Research Priorities and Challenges
selection in the kind of environment for which a variety is bred^ rather
than those of breeding stations. There is an additional reason for
carrying out selection in a poor environment. Thus, an experiment with
the fungus, Schizophyllum commune, in which selection was practiced for
both high and low growth rate at each of two different temperatures,
20° C and 30 °C, over several generations showed that, although they had
a virtually identical average growth rate, the high line selected at 20°C
(the poor environment) had a lower variance over temperatures than the
high line selected at 30°C (Jinks and Connolly, 1973). In addition to
displaying a greater stability, the mean growth rate of the high line
selected in the poor environment when raised at 30°C was not much less
than that of the high line selected at this temperature. Again, there is, no
reason for supposing that this outcome is peculiar to growth rate in
Schizophytlum. Data from the International Network for Genetic Evaluation
of Rice (INGER) could be used to establish that this relationship also
holds for rice. That varieties bred at Batalagoda, an environment which
for irrigated rice is not as good as that at Los Baños, have in general, a
wider adaptability than those bred at IRRI suggests that this is likely to
be the case.
The practical implications of the results from these experiments
with Schizophyllum and Nicotiana are clear, namely, that the initial
selection for high-yielding and stable genotypes in rice should be done
in poor environments and that subsequent testing of their stability
should be carried out in thé range of environments in which the crop is
grown by farmers. The question of whether the initial selection should
be carried out in farms or in poor environments simulated in breeding
stations is for those concerned to decide.
!■:
I
HYBRID VARIETIES
Discussion has been confined so far to breeding programs whose end
products are recombinant inbred lines. Given the apparent success of
hybrid varieties in China (Lin and Yuan, 1980; Yuan and Virmani, 1988),
a number of hybrid rice-breeding programs have been initiated
elsewhere, including IRRI. Hybrid varieties are, of course, bred to
exploit heterosis. There is no doubt that the Fj progeny of single crosses
of rice frequently display better-parent heterosis for yield; each of the
five crosses investigated by Perera et al. (1997) in which grain yield was
recorded (Crosses 6 - 1 0 ), for example, displayed better-parent heterosis
for this character. The breeding of hybrid varieties can be justified on
genetical grounds, however, if and only if, the genes for which the
parents differ display overdominance. If this is the case, no inbred line

M.J. Lawrence and D. Senadhira 137
extracted from a cross is expected to match the performance of the
original Fj hybrid; that is, the cross is not expected to yield superior
transgressive segregants. But as we saw earlier, the predicted yield of
the best F3 family was greater than that of the in each of Crosses 8-10
shown in Table 7,3. Again, though only one of the best families of
Cross 7 shown in Table 7.5 had an average yield that matched that of the
Fi of this cross (108.6 g), all four such families exceeded the yield of the
Pj in Cross 6 (98.3 g). This evidence of transgressive segregation in these
crosses indicates that the chief cause of the heterosis exhibited by their
Fj progeny is the dispersion of genes that display incomplete dominance
in the increasing direction between the parents. These results, therefore,
which are, of course, similar to those obtained from a wide range of crop
plant species (Jinks, 1983; Kearsey and Pooni, 1992), provide no genetical
justification for the breeding of hybrid varieties in rice. The heterosis,
which crosses frequently show, can be more easily fixed in recombinant
inbred lines extracted from them, thus avoiding the difficulties and
expense of having to breed male sterility into the female parent of
hybrid varieties. As explained earlier, however, heterosis can serve the
very useful purpose of identifying those crosses in which the genes are
dispersed between the parents and hence those capable of yielding
transgressive segregants in later generations of the pedigree. The chief
justification for hybrid varieties is the commercial advantage to breeding
companies of being able to protect the inbred parents of these varieties,
so that farmers are obliged to purchase their seed from these companies
every year rather than save seed from their own crops.
DISCUSSION
Adoption of the procedures discussed in this chapter should allow
breeders to exploit the genetical variation for each character of interest
in their crosses more efficiently and systematically than hitherto. The
problem of how best to improve the crop for all characters
simultaneously, remains, however. As seen above, the empirical
evidence indicates that the chief cause of the genetic correlation between
the majority of characters is the linkage disequilibrium of genes that are
linked in their inheritance, because the magnitude and sometimes also
the sign of these correlations, varies considerably over crosses. This
suggests that it should, in principle, be possible to choose crosses at the
F3 stage of pedigrees not only in terms of their potential to yield superior
transgressive segregants with respect each character considered
independently, but also those in which the pattern of genetic correlations
are most favorable to the breeders objectives in effecting improvement
for all characters simultaneously. Choosing crosses on this basis.

138 Rice Breeding and Genetics: Research Priorities and Challenges
however, would be justifiable only if the estimates of genetic correlation
obtained from the data of F3 families are sufficiently precise.
The question of the precision of estimates of genetic correlation
obtained frpm the generations of the selling series appears to have
received little attention. Some insight into this problem can be obtained,
however, from the modification of formulae developed for use with
data obtained from populations of individuals that mate at random
(Falconer and Mackay, 1996). Preliminary calculations showed that the
standard errors of estimates of genetic correlations between characters
obtained from F3 data sets of the size that Sriyoheswaran (1995) was
concerned with (40 families of size 10) can be uncomfortably large
unless the heritabilities of the characters are appreciable or the genetic
correlation between them is large or, of course, both. This is an
additional reason for confining attention to estimates of genetic
correlation whose absolute value is greater than 0.5. It would be prudent,
therefore, to disregard estimates when these conditions are not fulfilled.
However, this is only a first and partial step towards abatement of
this problem. What can be done to improve efficiency when selection is
carried out at the third, Fg stage of the biometrical breeding strategy? At
present, most breeders appear to use what Simmonds (1979) has referred
to as an ad hoc version of independent culling levels. In principle, the
most efficient way of tackling the problem would be to Omploy some
kind of index selection method, which has the great advantage of
combining information from the components of aggregate phenotype in
a systematic and objective way. While the simplest index is weight free,
most involve phenotypic and genetic variances and covariances of
characters and the same take into accoimt the economic value of each
component in the aggregate phenotype (see Baker, 1986 for a
comprehensive account of these methods). Although the economic
weights attached to components would presumably depend upon the
objectives of a breeding program and hence apply to all crosses in a
program, the phenotypic and genetic variances and covariances of
characters are, for reasons given above, expected to vary over crosses.
Thus, it would be necessary to compute any index that uses this
information for each cross separately. The use of an index of this kind,
however, would be worthwhile only if the estimates of these statistics
were sufficiently precise. Reassuringly, preliminary calculations of a
similar type to those used with estimates obtained horn the F3 families
show that estimates of genetic correlations between characters obtained
from Fg trials involving 1 0 0 families of size ten are very much more
precise than the former. This must also be true of the estimates of genetic
variances and covariances on which these genetic correlations are based.
Thus, although further work on this problem would be desirable, the

M.J. Lawrence and D. Senadhira 139
foregoing suggests that it might be worthwhile using a selection index to
identify those recombinant inbred lines extracted from a cross whose
performance comes closest to achieving the desired level for all
characters of interest to the breeder simultaneously.
Acknowledgements
We are indebted to our colleagues, Drs. M J. Kearsey and H.S. Pooni, for
discussion and comment on many aspects of the work presented here,
and to Professor RJ. Baker of the University of Saskatoon, Canada, for
help with the question of precision of estimates of genetic correlation.
References
Baker, R. 1986. Selection Indices in Plant Breeding. CRC Press, Boca Raton, Florida, USA.
Bentota, A.P., Senadhira, D. and Lawrence, M.J. 1997. Quantitative genetics of rice, III. The
potential of pair of new plant type crosses. Field Crops Res. (in press).
Boerma, H.R. and Cooper, R.L, 1975. Comparison of three selection procedures for yield in
soybean. Crop Sei, 15; 225-229.
Boughey, H. and Jinks, J.L. 1978. Joint selection for both extremes of mean performance and
of sensitivity to a macro-environmental variable. III. The determinants of sensitivity.
Heredity 40: 363-369.
Boughey, H., Jinks, J.L., Coombs, D.T. and Shufflebotham, W. 1978. Joint selection for both
extremes of mean performance and of sensitivity to a macro-environmental variable, IV.
Growth pattern and sensitivity. Heredity 41:175-183.
Brumpton, R.J., Boughey, H. and Jinks, J.L. 1977. Joint selection for both extremes of mean
performance and of sensivity to a macro-environmental variable, I. Family selection.
Heredity 38: 219-226.
Casali, V.W.D. and Tigchelaar, E.C. 1975a. Breeding progress in tomato with pedigree
selection and single seed descent, /, Amer. Soc. Hort. Sei. 100:362-364.
Casali, V.W.D. and Tigchelaar, E.C. 1975b. Computer simulation studies comparing
pedigree, bulk and single seed descent selection in self pollinated crops. /. Amer. Soc.
Hort. Sei. 100:364-367.
Cassman, K.G., De Datta, S.K., Oik, D.C., Alcantara, J.A., Samson, M., Descalsota, J. and
Dizon, M. 1994, Yield decline and the nitrogen economy of long-term experiments on
continuous, irrigated rice systems in the tropics. In: Experimental Basis of Sustainability and
Environmental Quality, R. Lai and B.A. Stewart (eds.). Lewis/CRC Publishers, Boca Raton,
Florida, USA, pp, 181-222.
Cornish, M.A. 1990a, Selection during a selling programme, I, The effects of a single round of
selection. Heredity 65: 201-211.
Cornish, M.A. 1990b. Selection during a selling programme, II. The effects of two or more
rounds of selection Heredity, 65:213-220,
Empig, L.T. and Fehr, W.R. 1971. Evaluation of methods for generation advance in a bulk
hybrid soybean population. Crop Sei. 11:51-54.
Fahim, M., Dhanapala, M.P., Senadhira, D. and Lawrence, M.J. 1997. Quantitative genetics
of rice, II. A comparison of the efficiency of four breeding methods. Field Crops Res. (in
press).

140 Rice Breeding and Genetics: Research Priorities and Challenges
Falconer, D.S. and Mackay, T.F.C. 1996. Introduction to Quantitative Genetics, Longman,
Harlow, Essex, UK (4th ed.).
FA O 1988. FAO World Crop and Livestock.Statistics 1948-1955. FAO Quart. Bull. Statistics, vol.
4, Rome.
Flinn, J.C., De Datta, S.K. and Labadan, E. 1982. An analysis of long-term rice yields in a
wetland soil. Field Crops Res., 5; 201-206.
Gravois, K.A., Moldenhauer, K.A.K. and Rohman, P.C. 1990. Genotype-by-environment
interaction for rice yield and identification of stable high-yielding genotypes. In: Genotype
by Environment Interaction and Plant Breeding, M.S. Kang (ed.). Louisiana State University,
Baton Rouge, LA, USA, pp. 181-188.
IRRI1989. IRRI Toward 2000 and Beyond. IRRI, Los Baños, Philippines.
IRRI1993. Rice Almanac 1993-1995. IRRI, Los Baños, Philippines,
Jinks, J.L. 1983. Biometrical genetics of heterosis. In: Heterosis, Monographs on Theoretical and
Applied Genetics, R, Frankel (ed.), Springer-Verlag, Berlin, pp. 1-46.
Jinks, J.L. and Connolly, V. 1973. Selection for specific and general response to environmental
differences. Heredity 30: 33-40.
Jinks, J.L. and Pooni, H.S. 1976. Predicting the properties of recombinant inbred lines derived
by single seed descent. Heredity 36:253-266.
Jinks, J.L. and Pooni, H.S. 1980. Comparing predictions of mean performance and
environmental sensitivity of recombinant inbred lines based upon F3 and triple test cross
families. Heredity 45; 305-312.
Jinks, J.L. and Pooni, H.S. 1987. The genetical basis of environmental sensitivity. In: Proc. 2nd.
Inti. Conf. Quantitative Genetics. B.S. Weireí al. (eds.), Raleigh, NC, USA.
Jinks, J.L., Jayasekara, N.E.M. and Boughey, H, 1977. Joint selection for both extremes of
mean performance and of sensitivity to a macro-environmental variable, II. Single seed
descent. Heredity 39j 345-355,
Kang, M.S. 1988. A rank-sum method for selecting high-yielding, stable corn gentoypes.
Cereal Res. Communi. 16:113-115.
Kaufmann, M.L. 1971. The random method of oat breeding for productivity. Can. J. Plant Sei.
15:13-16.
Kearsey, M.J. and Jinks, J.L. 1968. A general method of detecting additive, dominance and
epistatic variation for metrical traits, I. Theory. Heredity 23:403-409.
Kearsey, M.J. and Pooni, H.S. 1992. The potential of inbred lines in the presence of heterosis.
In: Reproductive Biology and Plant Breeding. Y. Dattee, C. Dumas and A. Gallais (eds.).
Springer-Verlag, London, UK, pp. 371-^385.
Kearsey, M.J. and Pooni, H.S. 1996. The Genetical Analysis of Quantitative Traits. Champman &
Hall, London, UK.
Khush, G.S., Brar, D.S., Bermett, J. and Virmani, S.S. 1994. Apomixis for rice improvement. In:
Apomixis: Exploiting Hybrid Vigor in Rice. G.S. Khush (ed.), IRRI Los Baños, Philippines,
pp. 1- 2 1 .
Knott, D.R. and Kumar, J. 1975. Comparison of early generation yield testing and single seed
descent procedure in wheat breeding. Crop Sei. 15:295-299.
Kropff, M.J., Peng, S., Setter, T.L., Matthews, R.B. and Cassman, K.G. 1994. Quantitative
understanding of rice yield potential. In: Breaking the Yield Barrier. K.G. Cassman (ed.).
Proc. Workshop on Rice Yield Potential in Favorable Environments. IRRI, Philippines, pp. 5-
20.
Lawrence, M.J. and Senadhira, D. 1997. Quantitative genetics of rice, IV. A breeding strategy.
Field Crops Res. (inpress).

M J. Lawrence and D. Senadhira 141
Lawrence, M.J. and Senadhira, D. 1998. A biometrical breeding procedure. IRRI Discussion
paper Series No.?? IRRI, Manila, Philippines (in press).
Lin, S.C. and Yuan, L.P. 1980. Hybrid rice breeding in China. In: Innovative Approaches to Rice
Breeding, IRRI, Los Baños, Philippines pp, 35-51.
Oeveren, A.J. van 1992. A comparison between single seed descent and early cross selection
in wheat breeding. Euphytica 58:275-287.
Oeveren, A.P, van and Stam, P. 1992. Comparative simulation studies on the effects of
selection for quantitative traits in autogamous crops: early selection versus single seed
descent. Heredity, 69: 342-351.
Park, S.J., Walsh, E.J., Reinbergs, E., Song, L.S.P. and Kasha, K.J. 1976. Field performance of
doubled haploid barley lines in comparison with lines developed by pedigree and single
seed descent methods. Can. J. Plant Sci. 56; 467-474.
Peng, S., Khush, G.S. and Cassman, K.G. 1994. Evolution of new plant ideotype for increased
rice potential. In: Breaking the Yield Barrier. K,G. Cassman (ed.), Proc. Workshop on Rice
Yield Potential in Favorable Environments. IRRI, Los Baños, Philippines, pp. 5-20.
Perera, A.L.T., Pahim, M., Sriyoheswaran S,, Dhanapala, M.P., Senadhira, D. and Lawrence,
M.J. 1997. Quantitative genetics of rice, I. Evidence of unexploited genetical variation for
yield and other quantitative characters in modern índica cultivars. Field Crops Res. (in
press).
Shukla, G.K. 1972. Some statistical aspects of partitioning genotype-environmental
components of variability. Heredity 29:237-245.
Simmonds, N.W. 1979. Principles of Crop Improvement. Longman, London and New York.
Sriyoheswaran, S. 1995. Quantitative inheritance of some agronomic characters of three wide
crosses in rice. Ph. D. thesis, Univ. Birmingham, UK.
Tee, T.W. and Qualset, C .0 .1975. Bulk populations in wheat breeding: comparison of single
seed descent and random bulk methods. Euphytica 24:393-405.
Thomas, W.T.B. and Tapsell, C.R. 1985. Cross prediction studies in spring barley, 3.
Correlation between characters. Theor, Appl. Genet. 71:550-555.
Yuan, L.P. and Virmani, S.S. 1988. Status of hyrbid rice research and development. In: Hybrid
Rice. IRRI, Los Baños, Philippines, pp. 7-24.

8
Insect and Disease
Resistance in Rice
A.P.K. Reddy* and J.S. Bentur*
INTRODUCTION
With the advent of the green revolution in rice during the late sixties^,
there was considerable, accentuation of insect and disease problems in
India. Although the extent of losses varied from region to region and
depended on several factors, instances of complete loss of the crop due
to one or two pests have often been reported. The most successful
strategy during the last two decades to manage insects and diseases has
been to grow resistant varieties and need-based application of
pesticides. Pest-resistant varieties are advantageous because their use
involves no additional cost nor knowledge base. Resistant varieties are
known for their compatibility with other methods such as biocontrol
and pesticides and are ecologically safe and socially acceptable. Further,
effectiveness of resistant plants is not affected by weather vagaries.
Considerable area under rice is currently protected from insect and
disease damage solely by host plant resistance, since viral and a few
bacterial diseases have no effective chemical protection. The information
available on the changing insect and disease pest scenario in India and
the efforts are underway to breed pest-resistant varieties to mitigate
losses associated with major pests are reviewed.
Directorate of Rice Research, Rajendranagar, Hyderabad 500030, India.

144 Rice Breeding and Genetics: Research Priorities and Challenges
MAJOR INSECT PESTS AND DISEASES
Among the 100 insect pests that affect rice crop, stem borers, in
particular the yellow stem borer (YSB) Scirpophaga incertulas, are the
most serious pests in India. Gall midge (GM), Orseolia oryzae causing
silver shoot damage comes next in importance. Among the planthoppers,
the brown planthopper (BPH) Nilaparvata lugens and the whitebacked
planthopper (WBPH) Sogatella furcifera are the most important
species, sometimes causing, total devastation of crop the green
leafhopper (GLH) Nephotettix viriscens and Nephotettix nigropictus cause
low damage to the crop as direct feeders but the former as a vector of the
tungro disease causes considerable losses indirectly. Rice hispa,
Didadispa armígera, traditionally a sporadic pest, is appearing in many
areas as a major pest. At times the leaffolder, Cnaphalocrocis medinalis,
thrips and gundhi bug (ear bug) also become serious.
Among the 85 diseases that affect rice crop, blast disease caused by
Pyricutaria grísea is considered a major production constraint. It affects
leaves, nodes, and panicles. Bacterial blight (BB), caused by Xanthomonas
oryzae pv oryzae, is a major problem in irrigated and rainfed ecosystems.
Sheath blight, caused by Rhizoctonia sotaní, assumed economic
importance during the 1980's because of changed management practices
and altered crop canopy structure. Sheath rots, hither to considered
unimportant as a group have lately become problematic in uplands and
rainfed lowlands of eastern India. Tungro disease continues to occur in
cyclic form and is causing substantial losses. False smut disease, caused
by Ustilaginoidea vireps, is gaining importance in northwest and eastern
India.
CHANGING INSECT AND DISEASE SCENARIO
i : ,1
During 1965-1995 the number of insect pests under "major pest status"
rose from 3 to 13 and in the case of diseases, from 2 to 8 . Planthoppers
and leafhoppers assumed major pest status with the advent of the green
revolution. The gall midge has become a major endemic pest in many
new areas. Sporadic pests such as hispa and the gundhi bug have been
causing serious losses in certain years. Yield losses due to insect pests
have been estimated to range from 21-30% (Kalode and Krishnaiah,
1991). Among diseases, recurrent crop losses were reported due to
epidemics of bacterial blight and rice tungro virus during the era of
high-yielding varieties (HYV). Consequent to varietal shift in the
eighties, blast reemerged as a major problem, especially in rabi (dry
season) rice areas. Sheath blight has been a major concern in eastern

A.P.K. Reddy and J.S. Bentur 145
India and the southern state of Kerela. In several stages^ sheath ro t false
smuh leaf scald and grain discoloration gained economic importance. In
general, it is estimated that 1 0 % grain is lost in tropical'Asia due to
diseases (Reddy, 1993).
FACTORS RESPONSIBLE FOR PEST ACCENTUATION
Before the onset of the green revolution, rice was grown in India for
centuries with traditional tall varieties. They lodged when excess
organic manners were applied to the field. Therefore, farmers grew
them under low to moderate levels of organic manure application.
Farmers generally planted several local varieties with diverse genetic
background in a mosaic fashion during any one season. Thus low
nitrogen application and genetic diversity in the subsistence ricecropping
system kept pests and disease problems to a minimum.
However, there were occasional epidemics of blast, brown spot and low
but chronic losses due to stem borers. But with the advent of
monoculture of high-yielding varieties with a narrow genetic base,
excessive use of synthetic nitrogenous fertilizers, absence of crop
diversification, and double cropping in the rice-rice cropping system
accentuated insect and disease problems. Besides varietal shift, some
other factors that accentuated disease problems were (a) changes in crop
management practices and (b) improper pesticide use. Of these factors,
new photo insensitive, dwarf, nitrogen-responsive, profuse tillering,
and short duration varieties with different plant architecture proved to
be the most congenial hosts for pest development; survival from season
to season led to rapid spread of these pests over vast stretches of rice
areas. Cultivation of HYVs changed crop management practices. Rice
cultivation in irrigated areas became a profitable enterprise tempting
farmers to higher use of nitrogenous fertilizers and pesticides to realize
higher returns. Absence of crop diversification coupled with
monoculture of high-yielding quality rices and absence of crop discipline
(lack of synchronous planting) proved to be other important reasons for
the upsurge of many pests. Indiscriminate use of pesticides, particularly
insecticides, was a significant cause of outbreaks of such insect pests as
BPH during the seventies and eighties (Kalode and Krishnaiah, 1991).
Destruction of natural enemies was found to be the cause for such
outbreaks (Kenmore, 1980), Leaf folders have also been observed in
severe from wherever pyrethroids and organophosphorous compounds
were overused.

146 Rice Breeding and Genetics: Research Priorities and Challenges
HOST PLANT RESISTANCE
Concepts and Levels of Resistance
M■I
In 1968/ Van der Plank recognized two basic types of resistance on the
basis of genetic and epidemiological concepts namely: (i) monogenic,
complete, race-specific resistance (vertical); and (ii) polygenic, incomplete,
race non specific resistance (horizontal). Race-specific resistance is
qualitative in nature and often characterized by hypersensitivity in a
host genotype. Race nonspecific, quantitative resistance has also been
called field resistance, generalized resistance, rate-reducing resistance,
dilatory resistance etc. (Lee et ah, 1989).
In recent years, a few other terms related to resistance, i.e. partial
and durable resistance, have come into vogue. Partial resistance is
quantitative and may be equated to field resistance. Durable resistance
is a retrospective concept and is based on the longevity of resistance in a
given ecosystem irrespective of genetic control and magnitude. It
essentially describes the commercial utility and longevity of a particular
resistant variety.
S p e c if ic r e s is t a n c e
[1ii
Specific resistance can either be immunity or high level of resistance to
insects and pathogens. In neither allows the pathogen to sporulate/
insect pest to multiply nor allows plant injury under any known
conditions, but it is short lived in nature and is often known to break
down, with serious economic consequences. In Korea, the resistance of
the "Tongil" varieties for blast was effective for a 5-year period. In
Japan, the longevity of resistant blast varieties is 3 years (Marchetti and
Bonman, 1989). In Colombia, resistant varieties to blast released during
1969-1986 lasted only for a year or two before being overcome by
previously unidentified virulent races (Ahn and Mulekar, 1986). In
many instances of investigation complete resistance for blast was
monogenic (Kiyosawa, 1981; Marchetti et ah, 1987). Although racespecific
resistance conferred by a single gene or a combination of a few
genes is generally short lived, as anticipated, there are many instances of
such genes proving effective for long periods and remaining effective in
many coimtries, viz. Pi-zt, Pi-b, Pi~ta2 and others. Transfer of major
resistance genes into elite cultivars is normally easy and has been
utilized successfully in many rice improvement programs,
notwithstanding the breakdown by the new genotypes of pathogen
populations.
Erosion of monogenic resistance against BPH (conferred by Bphl
gene) in IR 26 within 2 years of its intensive cultivation in the Philippines

A.P.K, Reddy and J.S. Bentur 147
and Indonesia in the mid-1970s is a classical example of inherent
weakness of monogenic resistance against insect pests (Cohen et al.,
1997). Similarly^ monogenic resistance in rice against GM conferred by
the Gm2 gene in the resistant variety Phalguna led to the breakdown of
resistance consequent to the development of Biotype 4 in parts of
Andhra Pradesh and Maharastra (Bentur et al., 1987).
P a r t ia l r e s is t a n c e
Varieties that lack complete resistance but sustain only minor losses
from blast are referred to as partially resistant. This kind of resistance is
quantitative^ race specific (Toriyama; 1975) or race nonspecific (Ezuka^
1979; Yeh and Bonman, 1986). In some rice-growing environments
highly conducive to blast, partial resistance may not be sufficient to
control the disease. In such situations, breeders exploit more complete
resistance through gene pyramiding or superimposing complete
resistance onto varieties possessing partial resistance (Bonman and
Mackill, 1988).
Moderate resistance against insects is characterized by a moderate
level of insect mortality imder a no-choice setup and relatively less plant
damage. Resistance known against the yellow stem borer in cultivated
rice can best be recognized as moderate resistance. Likewise, the
resistance known against polyphagous pests such as the leaf folder,
thrips etc. are of moderate level. Polygenic resistance to insects is
moderate but is more stable than monogenic resistance. Resistance
against YSB, which is polygenic, has not been lost so far. Nevertheless,
the moderate level of resistance has a key role to play in pest management.
In addition to the above, a recent concept field tolerance has been
introduced. Here the cultivar displays low resistance or even
susceptibility when tested as a young plant or under a very rigorous/
no-choice setup. But at the advanced stage of growth/maturity, such
varieties tend to register relatively less damage than the other
susceptible cultivars. Some of the BPH donors, e.g. Utri Rajappan,
Triveni and Kancana, have displayed such resistance/tolerance (Panda
and Heinrichs, 1983).
D u r a b l e r e s is t a n c e -
Durable resistance is that which has remained effective for a long period
while a qultivar possessing it has been widely cultivated in an
environment favoring the disease (Johnson, 1981). Durable resistance is
retrospective rather than prospective. The definite longevity required
for durable resistance is not predetermined. Resistance is referred to as
durable if resistance of one variety remains effective even though

i-:
'!
148 Rice Breeding and Genetics: Research Priorities and Challenges
another succumbs to an epidemic. The mode of inheritance is not fully
understood and methods to evaluate it are yet to be developed. Only
when such questions are answered, will breeding programs for durable
resistance attain practical value (Lee et al., 1989).
Some varieties considered having durable blast resistance are: IR 36
in tropical Asia (Yeh and Bonman, 1986); Milyang 30, Milyang 42 in
Korea (Lee et al, 1989); lAC 2 5 ,1 AC 47, Moroberekan from. West Africa,
(Nottenghem, 1985); Zhen Shan 97 and Zhen Luon 13 in China (Lee et al.,
1989). For bacterial leaf blight in China, resistance in Nongken 58 and a
few varieties possessing Xa3 and Xa4 genes was considered durable and
in the Philippines varieties possessing Xa4 resistance gene are
considered durable (Lee et al., 1989; Bonman et al, 1992). No such
information exists on durable resistance to RTV and other diseases.
GENETICS OF RESISTANCE
Insects
Several studies have been done on genetic characterization of resistance
against major insect pests such as BPH, WBPH, GLH, gall midge, stem
borer etc., mainly in the Philippines, India, Japan, and China. Due to the
prevalence of distinct geographic populations in these cotmtries, the
genetic information obtained in one country need not always hold true
in other countries. This is Well documented as several resistant donors
reported in one country are found to be susceptible in other countries.
Generally, resistance to insect pests is simply inherited with one or two
dominant/recessive genes (Table 8.1). While the complementary action
and inhibitory nature of some of the genes were noted, cytoplasmic
influence or strong environmental interaction with the genotype to
influence the phenotype is not common.
I
I
i
PLANTHOPPERS AND LEAFHOPPERS
Studies at IRRI by Khush and his group have identified five dominant
and five recessive genes against BPH four dominant and two recessive
genes against WBPH, six dominant and two recessive and two yet
unidentified genes against GLH and three dominant genes against the
!z;igzag leafhopper (Brar and Khush, 1991). Of these Bphl, bpH2, Wbphl,
Glhl, Glh2, Glh3 and GlhS are not functional against the South Asian
population of pests. Recent studies in China and Japan have suggested
the presence of several new genes conferring resistance against BPH and
WBPH which need to be characterized in terms of allelic relationship
with the known genes. Genetic studies in India revealed two or three

n
A.P.K. Reddy and J.S. Bentur 149
gene involvement in conferring resistance to BPH in several instances,
whereas a single dominant or recessive gene was responsible for WBPH
resistance. Recent quantitative trait loci (QTL) mapping studies at IRRI
(Alam, 1997) showed that the resistance of IR64 to the Central Luzon
and IRRI BPH populations is mostly governed by a complex of minor
genes which conform to the durability of resistance in the variety.
G a l l m id g e a n d S t e m b o r e r
Table 8 .1
Genes conferring resistance against major pests of rice
Type cultivar/variety (Source)
(3)
Pest
(1)
Gene
(2 )
Brown planthopper. Bfhl Mudgo
Nilaparvata lugens bph2 ASD7
Bph3
Rathu Heenati
bph4
Babawee
bpb5
ARC10550
Bph6
Swarrwlatha
bph7
T12
bphS
Chin saba
Bph9
Balamawee
BphlO(t)
IR65482-4-136-2’2
(Brar and Khush, 1991)
White-backed planthopper, Wbphl N22
Sogatdlafurcifera Wbphl ARC10239
Wbph3
ADR52
wbpH4
Podiwi AB
WbphS
N'Diang Marie
Wbph6(t)
Giu-yi-Gu
(Brar and Khush, 1991)
Green leafhopper. Glhl Pankhari 203
Nephottetix virescens Glh2 ASD7
Glh3
IR8
glh4
Ptb8
Glh5
ASD8
GIH6
TAPL796
Glh7
Maddai Karuppan
glhS
DV85
(Brar and Khush, 1991)
Gall midge, Gml W1263
Orseolia oryzae Gtti2 Siam 29
Gm4(t)
Abhaya
Gm6(i) Dukong No. 1
(Chaudury et al, 1985,
Srivastav et ai, 1994,
Katiyar et aL, 1996)
Blast, Pi-a Jae Keum
Pyricularia grísea Pi-b Tjina
Pi-I
Doazi Chali
Contd

150 Rice Breeding and Genetics: Research Priorities and Challenges
Pest
(1)
Gene
(2 )
Type cultivar/variety (Source)
(3)
Pi~k
Yakei-ko
Pi-kh HR2 2
Pi~kp
Pusur
Pi-ks
To-to
Pi-ta
Oka-ine
Pi-zt
Co25
(Kiyosawa, 1977)
Bacterial blight. Xal Kogyoku
Xanthomonas oryzae Xal Tetep
pv. oryzae Xa3 Wase Aikoku
Xa4
TKM6
Xa5
DZ192
Xa7
DV85
Xa8
PI231129
XalO
CaS209
Xal3
BJl
Xa21
Oryza longistamimta
(Ogawa and Kush, 1989)
A single dominant gene generally confers resistance against the gall
midge. While two dominant genes have been designated as Gmt and
Gm2, conferring resistance in rice varieties W1263 and Phalguna
respectively^ (Chaudhary et. ah, 1985), another seven genes have been
tentatively designated. In view of the prevalence of at least six different
biotypes in the country, biotype specific resistance was studied. It was
found that separate genes confer resistance to a specific biotype and
despite the presence of many dominant genes in a cultivar, only one of
them will express in response to attack by a specific biotype (Reddy et
at., 1997). Against the stem borer, one single recessive or three
complementary genes have been reported in variety TKM 6 . The allelic
relationship of these genes is not well studied.
Diseases
B l a s t
. ;l!
Scattered genetic studies on blast resistance started before physiologic
races were generally recognized. Many such studies used specific races.
From these studies it appeared that resistant varieties carried one, two,
or occasionally three genes for resistance and most of these genes were
dominant (Table 8.1). Kiyosawa and his colleagues in Japan did the most
extensive studies during the early seventies. They identified 13
resistance genes; two were found in japónica varieties while the others
were from exotic cultivars. Kiyosawa (1972) proposed the gene-for-gene
concept and linkages for genes known up to that time. Recent studies
have provided a great deal of new information. Marchetti et aL (1987)
identified three recessive genes that confer resistance to US races IC-17,

A.P.K. Reddy and J.S. Bentur 151
IG-1 and IH-1 . At the International Rice Research Institute (IRRI)/ the
Philippines and the Central Rice Research Institute (CRRI)^ India, a few
of the traditional rice cultivars evaluated had one or two dominant
genes. It was also observed that one or two dominant genes (< biblio >)
controlled isolate-specific resistance. Similar results were reported in
West Africa and Latin America, where several dominant resistant genes
were reported (Nottenghem, 1985).
B a c t e r ia l l e a f b l ig h t
Reports on genetics of resistance to BB indicate that major genes confer
high levels of resistance, which were qualitative in nature. At IRRI,
using the Philippine races of the pathogen, 15 genes for BB resistances
were identified (Table 8.1). Xal, Xa2, Xa3, and XalO convey a high
degree of resistance to a few Japanese and Philippine races of BB
patiiogen, whereas Xa3, Xa4, Xa5, Xa8, Xal3, and Xa21 confer a high
degree of resistance to the South Asian population of BB pathogen. X«4
is the major resistance gene present in several of the rice cultivars
released for commercial cultivation in the Philippines, viz. IR 20 through
IR 72. Several cultivars possessing Xai are also widely cultivated in
southern China. Genetics of BB resistance has been extensively studied
in India and several dominant/recessive genes have been reported
against the local races of X. oryzae (Table 8.1). Under Indian conditions
the functional resistant genes are Xa3, Xa5 + Xa7, Xa8, Xal3 and Xa21
(DRR, 1992).
R ic e T u n g r o V ir u s (RTV)
Studies on gerietics of RTV resistance are few. Preliminary studies
conducted in earlier years at IRRI and in India showed resistance to be
dominant (IRRI, 1966; Shastri et al,, 1972) and governed by a single or
two or three genes. Seetharaman et al. (1976) showed that three genes are
involved in resistant parents Kataribhog and Kamod 253^ Shahjahan et
al. (1991) reported that varieties Utri Mearah, Kataribhog, and Pankhari
203 were resistant to RTSV. It was observed that resistance in Utri
Mearah is controlled by a single recessive gene while three
complementary genes govern resistance in Pankari 203. The author also
noted that restrictive multiplication of RTBV in Utri Mearah was a
polygenic character. Studies at IRRI have shown that RTV is a complex
disease and its genetic analysis is complicated. Resistances to green
leafhopper (GLH) segregate in breeding material and interfere with
assessment of resistance to RTV (Hibino et al., 1987). A single dominant
gene governed RTSV resistance when analysis was done independently
by ELISA and antibiosis experiments.

152 Rice Breeding and Genetics; Research Priorities and Challenges
MECHANISM OF RESISTANCE
Insects
M o r p h o l o g ic a l t r a it s
i.'ll
A general association between morphological and anatomical characters
of the plant (like height and thickness of stem/ leaf blade width and
texture, leaf sheath compactness, length of the elongated internode, etc.)
and insect resistance was reported by earlier workers, but none per se
was shown to be the real cause of resistance (Pathak, 1969). Leaf
pubescence, for instance, did not account for the large differences noted
in number of eggs laid by the striped stem borer on the resistant hairy
leaves of TKM 6 and the susceptible glabrous leaves of Rexoro. Likewise,
leaf hairiness or compactness of leaf sheath implicated in gall midge
resistance (Roy et al., 1969; Rao et ah, 1971) were later refuted to be the
cause of resistance (Sain and Kalode, 1994). Studies on the leaf folder
showed only a weak positive correlation between leaf width and leaf
damage among varieties and a negative correlation between trichome
density on leaf and egg-laying preference by C. medinalis (Dakshayani et
ah, 1993),
A n t ix e n o s is (N o n p r e f e r e n c e )
Give a choice, insects generally prefer susceptible verieties to resistant
ones for alighting, shelter and oviposition. This is mainly a behavioral
response with a possible involvement of semiochemicals. However,
preference may also result from gustatory-mediated responses. A higher
number of adults/nymphs settling on susceptible varieties than on
resistant varieties is generally seen in the case of leaf and planthoppers
(Kalode, 1983; Baqui, 1990; Mishra and Misra, 1991). Such a response is
more pronounced with time lag, thereby indicating involvement of
feeding stimuli. Antixenosis for oviposition is more discrete and is
mediated through a different set of cues (Thompson and Pellmyr, 1991).
This phenomenon is widely reported for leafhoppers and planthoppers
(Kalode, 1983; Mishra and Misra, 1991) the stem borer and rice hispa,
but not for the leaf folder (Dakshayani et ah, 1993) or gall midge (Sain
and Kalode, 1994).
A n t ib io s is
More widespread and distinct physiological resistance involves
antibiotic effects of the resistant plant on the insect pest. These manifest
as reduced nymphal/larval survival, poor growth and development.

A.P.K. Reddy and J.S. Bentur 153
lowered pupation rate^, weaker adults and low build-up of pest
population through generations (Ramaraju and Babu, 1990), though
hypersensitive reaction against insect pest attack is rare but not
uncommon (Fernandes, 1990). Some of the rice varieties resistant to the
gall midge display hypersensitive reaction with tissue necrosis at the
site of attack leading to insect mortality (Bentur and Kalode, 1996). This
type of resistance is also inducible against a virulent biotype through
prior infestation by an avirulent biotype . The possible role of phenols is
suspected.
T o l e r a n c e
In some rice varieties, tolerance is noted as the basis of resistance,
especially against planthoppers (Panda and Heinrichs, 1983). Here the
host plant exhibits an ability to grow and reproduce normally or repair
injury due to insect feeding to a marked degree in spite of supporting a
population approximately equal to that which would severely damage a
susceptible variety (Panda and Khush, 1995).
P l a n t b io c h e m ic a l s
Correlative studies on resistant and susceptible plant varieties, generally
suggest that the amount of major elements such as nitrogen, potash, and
silica present in the plant tissue influences growth and development of
stem borers and leaf folders (Pathak 1969; Ramachandran and Khan
1991; Sudhakar et aL, 1991). Feeding studies on BPH revealed that oxalic
acid and p-sitosterol acted as deterrents (Hopkins, 1991) while higher
amino acid content and lower phenols favored GLH performance
(Viswanathan arid Kalode, 1990). Steam distillates of resistant/
susceptible varieties have been shown to elicit various behavioral/
physiological effects on insects. These effects have been attributed to
undefined allelochemicals.
Diseases
Infection of plants by fungal and bacterial pathogens results in an
inducible defense response. This can include synthesis and accumulation
of phytoalexins, reinforcement of cell walls by deposition of callóse,
lignin and related phenols, and an increase in the activity of hydrolytic
enzymes such as chitinases and gluconases. Most of the information
available on the mechanism of resistance pertains to a few fungal and
bacterial diseases. Early works are mostly correlation studies—
correlation of silicon content with resistance, and the increased content
of nitrogenous compounds including amino adds, amines and soluble

154 Rice Breeding and Genetics; Research Priorities and Challenges
nitrogen, with susceptibility. More recent studies have dealt with hostparasite
interaction and more specifically with host cultivar and
pathogenic race.
B l a s t
Different substances in plants such as epicuticular waxes, free phenols
and cell wall bound phenoloxidases and phytoalexins were reported to
operate against the blast pathogen. However, none of these mechanisms
are universal in nature (Sridhar et al, 1990). Cuticular waxes of varieties
susceptible to blast favor development of a large number of appressoria
when compared with those of resistant varieties. Intense tissue
browning, a characteristic feature of resistant varieties, limits the growth
and sporulation of the pathogen. The biochemical processes associated
with tissue browning affect pathogen development. Cell walls of rice
leaf blades contain cinnamate derivatives and the toxic activities of p-
coumarate, ferulate and their oxidized products are associated with
resistance. A few compounds such as probenazol were developed for
controlling blast disease, which augment development of phytoalexins
that reduce fungal growth. However, it was also shown that resistance
to P. grísea is not always governed by phytoalexins.
B a c t e r ia l l e a f b l ig h t
The resistance mechanism of rice cultivars to X. oryzae involves induced
resistance, protective reaction of antibacterial compounds and a few
other general mechanisms associated with plant defense mechanisms.
Induced resistance was noted when the rice plant was challenged with
an incompatible isolate of the pathogen prior to infection by a compatible
isolate. Also, preinoculation of either incompatible pathogen or
nonpathogen can induce resistance in rice against primary compatible
challengers. In some resistant varieties, ultrastructural changes of rice
plants to incompatible isolates of X. oryzae have been demonstrated
where the bacterial cells are immobilized by fibrillate material induced
from the cell walls of the host. A few workers have also demonstrated
the postinfection defense mechanisms with tissue specificity. A group of
antibacterial compounds such as trans 2 -hexenal, trans 2 -hexeonoic acid
and cos 3-hexeonic acid, syringaldéhyde, coniferaldehyde have been
identified and associated with resistance in rice plants to X. oryzae.
Purushothaman (1975) reported the presence of a larger amount of total
and ortho-dihydroxy phenols in resistant cultivars than in susceptible
cultivars. Reddy et al. (1977) suggested the role of phenols in the
restriction of BB pathogen in the host tissue.

A.PX. Reddy and J.S. Bentur 155
B r o w n s p o t
Some anatomical features of rice leaves have been observed to be related
to resistance. Several workers in earlier years have shown that thicker
epidermal cells and more silicated cells are positively connected with
resistance. Some reports also indicate the involvement of phenol
compounds and their oxidation system in the resistance mechanism
(Ou, 1985).
BREEDING FOR RESISTANCE AND GENETIC GAINS
Resistance to Insects
Resistant varieties play a vital role in the management of insect pests, in
particular the endemic ones, e.g. brown planthopper and gall midge.
Diversified sources of resistance to insect pests have been identified
through glasshouse and field screening, A total of 261 donors resistant
to BPH, 184 to WBPH, 194 to GM and 28 to leaf folder were identified.
Strong breeding programs supported by multilocation testing under the
coordinated program resulted in the development and release of 36
varieties resistant to gall midge, 24 to brown planthopper, 3 each to stem
borer and green leafhopper, and one to the white-backed planthopper
(Table 8.2). Of the GM resistant varieties, all are resistant to GM Biotype
1; 24 against Bio type 2; 11 against Bio type 3; 9 against Biotype 4; and 6
against Biotype 5. Many of these resistant varieties possessing high yield
and other desirable agronomic traits have been cultivated extensively in
pest-prone areas either as a principal method of control or as a
supplement to other methods of insect pest management.
Multiple Resistance against Insect Pests
Pest problems have become more complex in recent years because many
pests occur in a given area at the same time and cause significant damage.
Hence screening and breeding programs have been reoriented to develop
varieties with resistance to more than one insect pest. A number of donors,
such as Velluthacheera, ADR 52, Pandi, Chennellu, etc., have been
identified to possess multiple pest resistance (Kalode et aU, 1977).
Varieties developed with multiple pest resistance are listed below.
Variety
Suraksha
Shaktiman
Lalat
Rasmi
Daya
Samalei
Pests
GM, BPH, WBPH
GM, BPH, WBPH
GM, BPH,GLH
GM,BPH
GM,BPH,GLH
GM,BPH,GLH

156 Rice Breeding and Genetics: Research Priorities and Challenges
Resistance to Diseases
Host plant resistance has also been an important component in rice
disease management. Massive screening programs operated at various
levels in the country during the past two decades have identified several
resistant donors^ leading to the development of several commercial
resistant varieties for major diseases (Table 8.2). These varieties have
been utilized as a principal method for combating the pests.
B l a s t
ill
A large number of blast-resistant cultivars are available in India and
elsewhere. But a population of P, grísea quickly adapts to these varieties
and resistance breaks down. For instance/ the rice varieties NLR 9672/
Intan, Tellahamsa and others are no longer effective against the local
races of the pathogen. On the other hand/ many varieties grown in
rainfed lowland and irrigated rices do possess tolerance to blast disease.
A few widely cultivated blast-resistant cultivars are Rasb IR 36/
Swarnadhan among others. The primary donors for resistant cultivars
in Indian conditions have been Co4/ MTU5/ Zenith/ Tetep, Tadiikan, and
others. To avoid the boom-and-bust cycle, efforts are underway to
develop varieties that have better levels of quantitative resistance
(Reddy, 1993).
B a c t e r ia l l e a f b l iC h t
Of the 535 varieties released in India, about 35 possess a moderate level
of BB resistance. Only a few of these, viz. IR 20, IR 36, Saket 4, Swarna,
Mahsuri derivatives, Biraj, Radha, Sura], and Daya are widely grown. A
highly resistant variety, Ajaya, was recently released for commercial
cultivation (Reddy, 1993). Although a high degree of resistance to BB is
achieved through breeding programs, varieties with field tolerance
helped to reduce BB epidemics. Cultivars such as Saket 4, IR 20, Swarna
did not suffer much pest damage even when planted on large acres in
different states.
R ic e T u n g r o V ir u s
Breeding for resistance to RTV, as mentioned earlier, is complicated by
several factors. However, success has been achieved in developing RTV
resistant cultivars such as Vikramarya, Radha, lET 9994, and a few
others. Several new techniques coming into vogue, such as ELISA, could
simplify screening procedures for detection of RTSV and RTBV
independently and help to develop RTV-resistant varieties.

A.P.ÍC. Reddy and J.S. Bentur 157
Table 8,2
Resistance donors and varieties ivith resistance to major pests
developed and released for commercial cultivation in India.
Pest
Donors
(1)
(2 )
Brown planthopper. ARC 5984, ARC 6650,
Karivennel, Leb Mue
Nhang,
Manoharsali, Oorapundy,
Ptb 10, Ptb 18,
Ptb 21, Ptb 33
White-backed
planthopper Ptb 33 HKR120
Varieties released
(3)
Chaitanya, Krishnaveni,
Vajram, Pratibha,
Makom, Pavizham,
Manasarovar, Co42,
Chandana, Nagarajuna,
Sonasali, Rasmi,
Jyothi, Bhadra,
Neela, Annanga,
Daya, Aruna, Kanakaa,
Remya, Bharatidasan,
Karthika
Gall midge
C R 143; Eswarkora,
Leuang 152, Ob 677,
Ptb 21, Siam 29
Ptb 10, Ptb 18,
Divya, Dhanya Lakshmi,
Kakatiya, Erramallelu,
Kama, Ruchi, Orugallu,
Kavya, Rajendradhan 202,
Mdu 3, Buhban, Samalei,
Phalguna, Mahaveer,
Vibhava,
Pratap, Udaya, IR36,
Sarasa, Neela, Lalat,
Shakti, Suraksha, Daya,
Shaktiman, Tara, Kshira,
Sneha, Pothana,
Surekha, Vikram, Kunti
Usha, Asha, Abhaya
Stem borer TKM 6 Ratna, Sasyasree, Vikas
Blast
Bacterial Blight
Tetep, Tadukan,
Zenith, Co4,
Moroberekan, Correon,
Dissi Hatif, Taride 1
lAC 25, IRAT3
BJ1,TKM6,
Lacrosee, Zenith,
Nira, Java 14,
Wase-aikoku
Rasi, Akasi,
Sasyasree, Vani
Improved Sona, Morth 18,
Himdhan, Himalaya-1,
Himalaya-2, K332,
K333, HPU 741,
VLB, VLK 39,
NLR 9672, IRS,
IR20, IR36, IR64,
Pant Dhan 10, VL Dhan 221
Mahsuri, Prasad,
Ramakrishna, Saket 4,
Sasyasree, IET4141,
CNM540,IR20,
IR54, IR64,
Ajaya, Asha Daya
{Contd.)

158 Rice Breeding and Genetics: Research Priorities and Challenges
Pest
(1)
Sheath blight
Donors
(2)
T141, OS4
BCP3, Saibham,
Bhuhjan, Saduwee,
Laka, Ramedja,
Ta-Oo-Cho-z, Athebu
Phourel, ARC 15368
Varieties released
(3)
Pankaj, Swamadhan,.
Manasarovar
Brown spot
T141, BAMIO,
Chl3, Ch45
Rice tungro Ftb 2, Ptb 18, ADT 21,
ARC 10599, ARC 14320,
ARC 14766
Rasi, Jagannath
IR36, IR42
CNM529, CNM'540
Vikramarya, Radha
ïl :
RICE SHEATH BLIGHT
Most of the released varieties are highly susceptible to sheath blight and
a high degfee of resistance in O. saliva is not available. A few cultivars,
e.g. Pankaj and Swarnadhan^ have tolerance to this disease.
M u l t ip l e d is e a s e r e s is t a n c e
A rice crop in a given area may be vulnerable to attack by different
pathogens at different growth stages of the crop or^ due to the incidence
of certain diseases, be predisposed to other diseases. For example, RTV
or BB infections in the early growth stages weaken the plants and
predispose them to sheath rot infections. The synergistic action of two or
more diseases may cause extensive crop losses, indicating the need to
develop resistance for more than one disease (Reddy et al., 1986). But
until recently, no systematic efforts made to develop multirésistant
varieties, because most multiple disease-resistant donors have not been
available for commercial exploitation. The All-Ind|.a Coordinated Rice
Improvement Program (AICRIP) therefore undertook evaluation of
advanced breeding lines for multiple stresses under various
environments. A few commercial varieties developed for resistance/
tolerance to more than one disease are listed below.
Variety
Disease
IR36 Blast, Brown spot 1
Rasi Blast, Brown spot * ^
Vikramarya
RTV, Blast
Swamadhan
Blast, Sheath blight
Pankaj Sheath blight/Blast -1
Radha
Blast, Sheath blight
CNM539 Blast, Brown spot, and RTV %
In view of the complexity and high rate of mutability and evolution
of new pathotypes/races, it is more difficult to breed for multiple
resistance in a given ecosystem. Moreover, more complex "R" genes in a

A.P.K. Reddy and J.S. Bentur 159
host system preclude the development of desirable agronomic traits
such as yield quality/ etc. This has been the experience of breeding
programs; nonetheless, the search for multiple resistance in commercial
cultivars continues.
RESISTANT VARIETIES IN INTEGRATED POST MANAG1ÇMENT
Breeding pest resistant rice cultivars is one of the primary aims of rice
improvement programs worldwide as varietal resistance can be the
major strategy in rice integrated pest management (IPM), Most rice
farmers in Asia have small landholdings and derive relatively low
income from rice production. Some of the limitations, for example,
prohibitive pesticide cost, lack of credit facilities to purchase the
pesticides, lack of knowledge and skill to use pesticides effectively, and
the concomitant harmful effects of pesticides compel IPM planners to
overwhelmingly depend upon the genetic resistance in rice for pest
control. Peshresistant varieties are advantageous because their use
involves neither additional cost nor a knowledge base. Resistant
varieties are also known for their compatibility with other control
methods viz. biocontrol and cultural practices, and thus are ecologically
safe and socially acceptable. Further performance of resistant plants is
not affected by weather vagaries.
In gall midge endemic districts in Telengana and the northern
coastal districts of Andhra Pradesh, cultivation of gall midge-resistant
varieties such as Surekha and Phalguna in over 70% of the rice area
brought down pests to a minor level, resulting in about 45% increase in
yield levels. Similar results were realized by cultivation of BPH-resistant
varieties Chaitanya, Vajram, Krishnaveni in the coastal districts of
Andhra Pradesh; M 05, M 06, M 07 and Jyothi in Kuttanad area of
Kerala. In stem borer prone areas, cultivation of moderately resistant
varieties Vikas and Sasyasree needed occasional additional protection
(Krishnaiah and Reddy, 1989; Desai, 1987). Vikramarya and lET 9444
could successfully check the viral disease (Reddy and Reddy, 1993).
USE OF BIOTECHNOLOGY
In the past, genetic improvement for pest resistance has been achieved
mainly through the application of classical genetics and conventional
plant-breeding methods. Plant breeders have relied upon the primary
gene pool. The growing complexity of the pest-disease syndrome
warranted newer and novel genes/means beyond the conventional gene
pool, however. Recent advances in the fields of cellular and molecular

I
I
160 Rice Breeding and Genetics: Research Priorities and Challenges
biology and available tools of genetic engineering offer an array of
innovative approaches to exploit rare and novel gene sources from
distantly related species and even unrelated organisms and move these
genes into the rice genome.
Cell/tissue Culture Techniques
Among various cell/tissue culture techniques, embryo rescue and
somacloning find extensive application in genetic enhancement of rice.
Using the embryo rescue technique resistance genes for BPH and WBPH
were successfully transferred to elite O. sativa cultures from O. officinalis,
O. minuta, O. latifolia, O, australiensis and O. granulata (Jena and Khush,
1987; Ye and Saxena, 1990; Velusamy, 1991; Brar and Khush, 1995). The
YSB resistance gene from O. hrachyantka is also being to transferred to
cultivated rice (Bennett et ah, 1997).
Molecular Markers
Marker-aided selection offers greater advantage for gene pyramiding.
Gene tags useful for BB resistance genes Xai, Xa5, Xal3, and Xa21 have
been identified and suitable selectable markers developed. These
markers help in the identification of two or more genes that have been
combined in one individual plant and assess the effectiveness of 2 - 4
gene pyramids (Zhang et ah, 1996; Huang et ah, 1997). Three gall midgeresistant
genes—Gm2, Gm4(t), Gm6(t) and one gene each for BPH-
BphW(t), QLU~~Glh? (in ARC 11554) and WFBH~~Wbphl have been
tagged and mapped (Bennett et ah, 1997). Once suitable markers are
available for these and many more similar genes, tasks like gene
pyramiding for durable and multiple resistance breeding will be made
easier.
Gene Cloning and Introgression of Cloned Genes
A range of novel genes for insect resistance is now available for
incorporation into any plant genome (Carozzi and Koziel, 1997). Genes
from the ubiquitous soil bacterium. Bacillus thuringiensis (Bt), encoding a
class of insecticidal crystal proteins, have been successfully incorporated
into the rice genome. Transgenic plants with crylAh (Fujimoto et ah,
1993; Wunn et ah, 1996; Ghareyazie et ah, 1997; Wu et ah, 1997; Cheng et
ah, 1998) and cry 1 Ac (Nayak et ah, 1997, Cheng et ah, 1998) genes have
been reported to be resistant to rice pests such as YSB, LF, and striped
stem borer. Genes for expression of protease inhibitors from plant

A.P.K. Reddy and ).S. Bentur 161
sources have also been used in rice transformation and these plants
carrying potato proteinase inhibitor II (Duan et al, 1996), cowpea trypsin
inhibitor (CpTi) (Xu et al, 1996) and corn cystatin (Irie et ah, 1996) genes
are also insect resistant. These and many more of the new resistance
genes now available will fill the gap in the primary gene pool; thus a
high level of resistance against such pests as YSB is now feasible.
Good sources of resistance against sheath blight fungus Rhizoctonia
solani are lacking. A novel gene encoding chitinase enzyme has been
incorporated into the rice genome using a constitutive promoter
CaMV35S (Datta et ah, 1996). These transgenic plants are currently being
evaluated for sheath blight resistance. Novel genes have also been
transgressed into rice to confer resistance against viral diseases.
Successful transformation include: against RTBV (rice tungro bacilliform
virus) a gene encoding complete or mutated viral protein (Kloti et ah,
Í996); against RTSV (rice tungro spherical virus) a gene for coat protein
(Shivamani et ah, 1996)^ and against RRSV (rice ragged stunt virus) basal,
genome segments coding nonstructural protein (Upadhyaya et al., 1996).
The new BB gene Xa21 has been cloned (Ronald et ah, 1994) and has been
used for transformation of several commercial varieties. Some of these
transgenic rices are proposed for field evaluation in the Philippines
during 1998 (Benneth pers. commu.). Use of such a native gene to
develop new resistant varieties through transformation technology does
not alter the rest of the genetic constitution of the variety transformed
and thus saves time involved in conventional breeding through several
cycles of backcrosses.
DNA Fingerprinting and Diversity in Pest Populations
The third area in which biotechnology aids utilization of host plant
resistance is understanding biodiversity in pest populations. A series of
repetitive DNA elements have been isolated from the genome of X.
oryzae, causal organism for BB, and used for DNA fingerprinting of
various collections of pathogen strains. Spatial distribution of the
pathogen within India was delineated and functional resistance genes
for the common lineages/pathotypes were determined (Yashitola et ah,
1997). Restricted fragment length polymorphism (RFLP) analysis has
been successfully used to detect genetic variability in Asia (Adhikari et
ah, 1995; George et ah, 1997). Considerable progress has been made in
the last decade in studying population variability of the blast pathogen
P. grísea. Sexually compatible strains of the fungus have been identified
in natural populations and molecular karyotypes have been determined.
Also, population biology of the fungus has been clarified and progress
toward genetic identification, cloning, and characterization of

162 Rice Breeding and Genetics: Research Priorities and Challenges
transposable elements and host/cultivar specificity genes has been
made. Information gained from these analyses played a key role in the
development of national breeding strategies such as lineage exclusion
(Leong et al, 1994; Zeigler et ah, 1994). DNA fingerprinting involving
APLP techniques has recently been used in a study on Asian rice gall
midge populations from six different countries (Bermett et ah, 1997).
This study suggested that evolution of the new gall midge biotype (now
designated as Biotype 6 ) in the northeastern state of Manipur (India)
was more likely through migration of the pest from China rather than
through selection pressure of the resistant varieties.
CONCLUSIONS
Utilization of commercial rice varieties possessing pest resistance for
effective and economic management of insect pests has helped to
stabilize rice production. Multiple pest damage and continuous
emergence of new pest or biotype/pathotype problems have been
posing new challenges to the ongoing crop improvement efforts. Further
efforts are needed to breed multiple pest-resistant varieties with
polygenic background to confer wide-range resistance. Ecological
understanding of the pest population structure helps in developing
suitable deployment strategies for achieving durable resistance. A large
proportion of the idéntified sources of resistance still remains unutilized.
Untapped resistant genes available from land races and wild accessions
are yet to be explored. Cellular and molecular techniques also offer a
wide range of novel means to enrich the existing resource base. Utilizing
pest-resistant varieties as the core^ area-specific IPM programs could be
developed. Thus there is ample scope for developing multiple-resistant
varieties with polygenic background to meet the new challenges of the
next millenium.
References
Adhikari, T.B., Veeracruz, C.M., Zhang, Q., Nelson, R.J., Skinner, D.Z., Mew, T.W. and Leech,
J,E. 1995. Genetic diversity of Xanthomonas oryzae pv. oryzae in Asia. App/. Environ.
Microbiol. 61; 966-971.
Ahn, S.W. and Mulekar, A, 1986. Rice blast management under upland conditions In:
Progress in Upland Rice Research. IRRI, Los Baños, Philippines, pp. 363-374.
Alam, S.N. 1997. Durability and genetic basis of resistance of rice variety IR64 to brown
pianthopper, Nilaparvata lugens (Stál). Ph.D. thesis, Univ. Philippines, Los Baños, Philippines,
Baqui, M.A, 1990. Varietal preference of rice brown pianthopper Nilaparavata lugens (Stál)
(Homoptera; Delphacidae). Bangladesh /. Zoo/. 18:119-121.

A.P.K. Reddy and J.S. Bentur 163
Bennett J-, Cohen, M,B., Katiyar, S.K. Ghareyazie, B. and Khush, G.S, 1997. Enhancing insect
resistnacein rice through biotechnology. In: Advance^ in Insect Control; Role ofTransgenic
N.B. Carozzi and M.G. Koziel, (eds.). Taylor and Francis, London, pp. 79-93.
Bentur, J.S. and Kalode, M.B. 1996. Hypersensitive reaction and induced resistance in rice
against Asian rice gall midge Orseolia oryzae. Entomol. Exp. Appl. 78: 77-81,
Bentur, M.B., Srinavasn, T.E. and Kalode, M.B, 1987. Occurrence of a virulent of rice gall
midge (GM) Orseolia oryzae Wood-Mason biotype (?) in Andhra Pradesh, India. Inti. Rice
Res. Newslett, 12; 33-34.
Bonman, J,M. and Mackill, D.J. 1988. Durable resistance in rice. Oryza 25:103-110.
Borunan, J.M., Khush, G.S, and Nelson, R.J. 1992. Breeding rice for resistance to pests. Ann.
Rev, Phytopathol. 30: 507-528..
Brar, D.S, and Khush, G.S. 1991. Genetics ofresistance to insects in crop plants. Advances
Agron. 45: 223-274.
Brar, D.S. and Khush, G.S. 1995. Wide hybridization for enhancing resistance to biotic and
abiotic stresses in rainfed lowland rice. Rroc. Inti. Rice Res. Conf. Feb. 13-17,1995.
Carozzi, N.B. and Koziel, M.G. 1997. Advances in Insect Control: Role of Transgenic Plants.
Taylor and Francis, London.
Chaudhary, B.F., Srivastava, P.S., Shrivastava, M.N. and Khush, G.S. 1985. Inheritance of
resistance of gall midge in some rice cultivara of rice. In: Rice Genetics. IRRI, Los Baños,
Philippines, pp. 523-527.
Cheng, X,, Sardana, S., Kaplan, H. and Altosaar, 1 .1998. Agrobacterium-transformed rice
plants expressing synthetic cryIA(b) and cryIA(c) genes are highly toxic to striped stem
borer and yellow stem borer. Proc. Natl. Acad. Sci. USA 95:2767-2772.
Cohen, M.B., Alam, S.N., Medina, E.B. and Bernal, C.C. 1997. Brownplanthopper,Ní7flpflttJ«ffl
lugens, resistance in rice cultivar IR64: mechanism and role in successful N. lugens
management in Central Luzon, Philippines. Entomol. Exp. Appl. 85:221-229.
Crill, J.P. and Khush, G.S. 1982. Effective and stable control of rice blast with monogenic
resistance. In: Evolution of the Gene Rotation Concept for Rice Blast Control. IRRI, Los Baños,
Philippines, pp. 87-102.
Dakshayani, K., Bentur, J.S. and Kalode, M.B. 1993. Nature of resistance in rice varieties
against leaffolder Cnaphalocrods medinalis (Guenee). Insect Sci. Applic. 14:107-114.
Datta, S.K., Quimo, C., Torrizo, L., Datta, K., Alam, M.F., Abrigo, E., Oilvia, N., Alejar, M.,
Biswas, S., Muthukrishnan, S., Potrykus, I., Sellar, T., Mew, M.W. and Khush, G.S. 1996.
Genetic engineering for resistance to sheath blight and other agronomic characters. In:
Rice Genetics III. IRRI Los Baños, Philippines, pp. 785-791.
Desai, B.D. 1987. Operational research projects: constraints and achievements. Annu. Rice
Workshop, Patna, India (mimeo).
Directorate of Rice Research (DRR) 1992. Progress report of the AICRIP Kharif 1992, vol. 2.
Directorate of Rice Research, Hyderabad, India.
Duan, X-, Li, X., Xuc, Q., Abo-El-Saad, Xu, D. and Wu, R. 1996. Transgenic rice plants
harboring an introduced potato proteinase inhibitor II gene are insect resistant. Nature
Biotechn. 14; 494-498.
Bzuka, A. 1979. Breeding for genetics of resistance in Japan. In: Proc. Rice Blast Workshop.
IRRI, Los Baños, Philippines, pp. 27-40,
Fernandes, G.W. 1990. Hypersensitivity: A neglected plant resistance mechanism against
insect herbivores. Environ. Entomol. 19; 1173-1182.
Fujimoto, H„ Itoh, K., Yamamoto, M., Kyozuka, J. and Shimamoto, K. 1993. Insect resistant
rice generated by modified delta endotoxin gene of Bacillus thuringiensis. Bio techn. 11:
1151-1155.

164 Rice Breeding and Genetics: Research Priorities and Challenges
George, M.L.C., Cruz, W.T., Leach, J.E. and Nelson, R.J. 1997. Movement of Xanthomonas
oryzae in Southeast Asia detected using PCR based DNA fingerprinting. Phytopath. 87:
302-309.
Ghareyazie, B., Alinia, P., Menguito, C.A., Rubia, L.G., dePamla, J.M., Liwanag, E.A., Cohen,
M. B., Khush, G.S. and Bennett, J. 1997. Enhanced resistance to two stem borers in an
aromatic rice containing a synthetic crylA(b) gene. Molec. Breed. 3:401-414.
Hibino, H., Tiongco, E.R., Cabinanagow, R.C. and Florez, Z.M. 1987. Resistance to rice
tungro associated viruses in rice under experimental and natural conditions. Phytopath.
77; 871-875.
Hopkins, R.M. 1991. Feeding behaviour of the brown planthopper (BPH) on susceptible and
resistant rice cultivars. Jnt. Rice Res. Newslett. 16{6): 10.
Huang, N., Angeles, E.R., Domingo, J., Magpantay, G,, Sing, S., Zhang, G., Kumara Vaditel,
N. , Bennettl, f, and Khush, G.S. 1997, Pyramiding of bacterial blight resistance genes in
rice; marker assisted selection using RFLP and PGR. Theor. Appl. Genet. 95:313-320.
International Rice Research Institute (IRRI) 1996. Annual Report for 1995. IRRI, Los Baños,
Philippines.
Irie, K-, Hoypyama, H., Takeuchi, T., Iwabuchi, K., Watanabe, H., Abe, M., Abe, K. and Arai,
S. 1996. Transgenic rice established to express com cystatin exhibits strong inhibitory
activity agáinst insect gut proteases. Plant Molec. Biol, 30:149-157.
Jena, K..K. and Khush, G.S. 1987. Introgression of genes fromOrt/za officinalis Wall et Watt, to
cultivated rice O. sativa. Theor. Appl. Genet. 80; 737-7^5.
Johnson, R. 1981. Durable resistance: Definition of genetic control and attainment in plant
breeding. Phytopath. 71; 567.
Kalode, M.B. 1983. Leafhopper and planthopper pests of rice in India. In: Inti. Workshop on
Leafhoppers and Planthoppers of Economic Importance. W.J. Kright, N.C. Pant, T.S. Robertson,
and M.R. Vilson (eds.). Commonwealth Institute of Entomology, London, pp. 225-245.
Kalode, M.B. and Krishnáiah, K. 1991. Integrated pest management in rice. Indian /. Plant
Protect. 19; 117-132,
Kalode, M.B., Trishna, T.S., Pophaly, D,J. and Laxminarayana, A. 1977, Note on new multiple
resistant donors to major insect pests of rice. Indian /. Agrie. Sei. 47(12) 626; 627.
Katiyar, S.K., Tan, Y., Zhang, Y., Huang, B,, Xu, Y., Zhao, L., Huang, N., Khush, G.S, atvd
Bennett, J. 1995. Molecular tagging of gall midge resistance genes in rice. In: Fragile Lives
in Fragile Ecosystems. IRRI Los Baños, Philippines, pp: 935-948.
Kenmore, P.E, 1980. Ecology and outbreaks of a tropical insect pest of the green revolutionthe
rice brown planthopper Nilaparvata lugens (Stäl). Ph.D. thesis, Univ. California,
Berkeley.
Kiyosawa, S. 1972. Genetics of blast resistance. In; Rice Breeding. IRRI, Los Baños, Philippines.
Kiyosawa, S. 1977. Genetics of blast resistance. In: Rice Breeding, IRRI, Los Baños, Philippines,
pp. 203-225.
Kiyosawa, S. 1981. Gene analysis for blast resistance. Oryza 18:196-203.
Kloti, A. J., Fuherer, R., Terada, S., Wurm, J,, Burkhardt, P.K., Chen, G., Höhn, Th., Ghosh, G.C.
and Potrykus, 1.1996, Toward genetically engineered resistance to tungro virus. In: Rice
. Genetics III, IRRI, Los Baños, Philippines, pp. 763-767.
Krishnaiah, K. and Reddy, P.C, 1989. Operational research project on integrated control of
rice pests in Medchal area of Ranga Reddy district of Andhra Pradesh. Ann. Rice
Workshop, Hissar, India (mimeo).
Lee, E.J,, Zhang, Q, and Mew, T.W, 1989. Durable resistance to rice diseases in irrigated
environments. In: Progress in Irrigated Rice Research. IRRI, Los Baños, Philippines, pp.
93-110.

A.P.K. Reddy and J,S. Bentur 165
Leong, S.A., Farman, M., Smith, J,, Budda, A., Tora, Y. and Nitla, 1994. Molecular genetic
approach to the study of cultivar specificity in the rice blast fungus. In; Rice B/ast Disease.
R. S. Zeigler, S.A. Leong, and P.S. Teng, (eds.). Los Baños, Philippines, pp. 87-110.
Marchetti, M.A. and Bonman, J.M. 1989. Rice blast disease management. In: Rice Farming
Systems: New Direciiofxs. IRRI, Los Baños, Philippines, pp. 175-183.
Marchetti, M.A., Lai, X.H. and Bollach, C.N. 1987. Inheritance of resistance to Pyricularia
oryzae in rice cultivars grown in United States, Phytopath. 77: 799-804.
Mishra, N.C. and Misra, B.C. 1991. Orientation of white backed planthopper, SogatdU
furcifera (Horvath) towards some rice varieties. Indian }. Plant Profeci. 19:179-181.
Nayak, P., Basu, D., Das, S., Basu, A., Ghosh, D., Ramakrishnan, N.A., Ghosh, M. and Sen,
S. K. 1997. Transgenic elite indica plants expressing CrylAc 5-endotoxin of BíícíVíms
thuringiensis are resistant against yellow stem borer (Scirpophaga incertuks). Proc. Natl.
Acad. Sci. USA, 94; 2111-2116.
Nottenghem, J.L. 1985. Blast resistance methodologies in the Ivory Coast, 1972-1985. In:
Progress in Upland Rice Research. IRRI, Los Baños, Philippines, pp. 305-316.
Ogawa, T. and Khush, G.S. 1989. Major genes to bacterial blight in rice. In: Bacterial Blight of
Rice. IRRI, Los Baños, Philippines, pp. 177-192.
Ou, S.H. 1985, Rice Diseases. Commonwealth Institute of Mycology, Kew, England.
Padmanabhan, S.Y. 1965. Breeding for blast resistance in India. In: The Rice Blast Disease of
Rice. John Hopkins Press, Baltimore, MD, USA, pp. 343-359.
Panda, N. and Heinrichs, E.A. 1983. Levels of tolerance and antibiosis in rice varieties
having moderate resistance to the brown planthopper, Nilaparvata lugens (Stál)
(Hemiptera: Delphacidae). Environ-. Entomol. 12; 1204-1214.
Panda, N. and Khush, G.S. 1995. Host Plant Resistance to Insects. IRRI, Los Baños, Philippines
and CAB International, Oxon, UK,
Pathak, M.D. 1969. Stem borer and leafhopper-planthopper resistance in rice varieties.
Entomol. Exp. Appl. 12: 789-800.
Purushothaman, D. 1975, Changes in phenolic compounds in rice varieties as influenced by
Xanthomanas oryzae infection. Rice, 25:85-89.
Ramachandran, R, and Khan, Z.R. 1991. Mechanisms of resistance in wild rice Oryz«
brachyantha to rice leaffolder Ciiap/w/ocracis medinalis (Guenee) (Lepidoptera: Pyralidae).
/. Chem. Ecol. 17:41-65.
Ramaraju, K. and Babu, P.C.S. 1990. Growth and development of whitebacked planthopper
{Sogatelkfurcifera) on rice (Oryza sativa). Indian /. Agrie. Sci. 60:249-251.
Rao, Y.S., Israel, P., Yadava, C.P. and Roy, J.K. 1971. Nature of gall midge resistance in rice,
Curr. Sci. 40:497-498.
Reddy, A.P.K. 1993. Changing scenario: rice diseases and their management. In: Pest and
Disease Management in India-The Changing Scenario. Plant Protection Assoc. India,
Hyderabad, India, pp. 121-125.
Reddy, A.P.K. and Reddy, P.C. 1993. Integrated disease management in rice. In: Plant
Protection and Environment. D.V.R. Reddy, H.C. Sharma, T.B. Gour, and B.J. Divakar,
(eds.). Plant Protection Assoc. India, Hyderabad, India, pp. 239-259.
Reddy, A.P.K., Miah, S.A. and Mew, T.W. 1986. Disease resistance in rainfed lowland rice.
In; Progress in Rainfed Lowland Rice. IRRI, Los Baños, Philippines, pp. 253-262.
Reddy, A.V., Rao, U.P., Siddiq, E.A. and Bentur, J,S, 1997 Genetics of resistance to rice gall
midge (Orseolia oryzae). Indian }. Genet Plant Breed. 57:361-372.
Reddy, P.R., Nayak P. and Sridhar, R. 1997. Physiology of bacterial leaf blight of rice:
Influence of light intensity on some biochemical changes associated with disease
development. Indian Phytopathol. 30:51-54.

Ti
166 Rice Breeding and Genetics: Research Priorities and'Challenges
/íl
Ronald, P.C., Holtein, T., Scambray, J., Song, W., Wang, G. and William, C. 1994. Molecular
genetic analysis of the rice bacterial blight resistance locus. Xa2I. In; Blast Disease of Rice.
R.S. Zeigler, S.A, Leong, and P.S. Teng, (eds.). Los Baños, Philippines, 187-194.
Roy, J.K., Israel, P. and Pan war, M.S. 1969. Breeding for insect resistance in rice. Oryza 6:38-
44.
Sain, M. and Kalode, M.B. 1994. Greenhouse evaluation of rice cultivars for resistance to gall
midge, Orseolia oryzae (Wood-Mason) and studies on mechanisms of resistance, insect Set.
Appl. 15: 67-74.
Seetharaman, R., Prasad, K. and Anjaneyulu, A. 1976. Inheritance of resistance to tungro
disease. Indian/. Genet. Plant Breed. 36:34-36.
Shahjahan, M., Imbe, T., Jalani, B.S., Zakri, A.H. and Othman, O. 1991. Inheritance of
resistance to rice tungro virus in rice {Oryza sativa L.). In: Rice Genetics II. IRRI, Los Baños,
Philippines, pp. 247-254,
Shastri, S.V.S., John, V.T. and Seshu, D.V. 1972. Breeding for resistance to rice tungro in
India. In: Rice Breeding. IRRI, Los Baños, Philippines, pp. 239-252,
Shivamani, E . , Shen, P., Ong, C .A., Hamidah, G., Hasan, G., Mohamad Senawi, M.T., Beachy,
R.N. and Fauquet, C.M, 1996. Biological screening and analysis of transgenic rice lines
expressing coat protein genes of rice tungro spherical virus. In: Rice Genetics III. IRRI,
Manila, Philippines, pp .768-772.
Sridhar, R., Nayak, M. and Kumar, S. 1990. Physiology of disease resistance in rice blast
fungus. In; Extended Summary-Proc. Inti. Symp. Rice Research; New Frontiers.
Directorate of Rice Research, Hyderabad, India, 221 pp,
Srivastava, M.N., Kumar, A., Shrivastava, S.K. and Sahu, R.K. 1994. A new gene for gall
midge resistance in rice. Rice Genet. Neivslett, 10: 79-80.
Sudhakar, G.K., Singh, R. and Mishra, S.B. 1991. Susceptibility of rice varieties of different
duration to rice leaffolder Cnaphalocrocis medinalis Guen. Evaluated under varied land
situations,/, Entomol,/Ies. 15: 79-87.
Thompson, J.N. and Pellmyr, O. 1991. Evolution of oviposition behaviour and host
preferences in Lepidoptera. Ann. Rev. Entomol, 36:65-89,
Toriyama, K, 1975. Recent progress of studies on horizontal resistance in rice breeding for
blast resistance in Japan. In; Horizontal Resisfimee to Blast Disease of Rice. CIAT, Cai,
Colombia, pp. 65-100,
Upadhyaya, N.M., Ram, K,, Yang, M., Kosiratno, W. and Waterhouse, P.M. 1996. Rice ragged
stunt virus synthetic resistance genes and japónica rice transformation. In: Rice Genetics
III, IRRI, Los Baños, Philippines, pp. 773-779.
Van der Plank, J.E. 1968. Disease Resistance in Plants. Acad. Press, NY.
Velusamy, R. 1991. Resistance of breeding lines derived from Oryza officinalis to brown
planthopper (BPH). iui. Rice Res. Neu/slell. 16(1): 14.
Viswanathan, K. and Kalode, M.B. 1990.. Biochemical aspects of varietal resistance to green
leafhoppers, Nephotettix virescens (Distant) and Nephotettix nigropictus (Stál). Proc. Indian
Acad. Sci., Anim. Sci, 99; 57-66,
Wu, C., Fan, Y., Zhang, C,, Oliva, N. and Datta, S.K. 1997. Transgenic fertile japónica rice
plants expressing a modified cry lA(b) gene resistant to yellow stem borer. Pianf Cell Rep.
17; 129-132.
Wünn, J.f Kloti, A., Burkhardt, P.K., Ghosh Biswas, G.C., Launis, K., Iglesias, V. and Potrykus,
I. 1996. Transgenic indica rice breeding line IR58 expressing a synthetic cryIA(b) gene
from Bacillus thuringlensis provide effective insect pest control. Bio tech. 14:171-176.
Xu, D., Xue, Q., McElory, D., Mawal, Y., Hilder, V.A. and Wu, R. 1996. Constitutive
expression of a cowpea trypsin inhibitor gene Cp Ti in transgenic rice plants confers
resistance to two major rice insect pests. Mol. Breed. 2:186-193.

A.P.K. Reddy and J.S. Bentur 167
Yashitola, I., Kkrishnaveni, D., Reddy, A.P.K. and Sonti, R. 1997. Genetic diversity within a
population of Xanthomonas oryzae pv oryzae in India. Phytopath. 87: 765-775.
Ye, Z.H. and Saxena, R.C. 1990. Resistance to whitebacked planthopper in elite lines of
cultivated x wild rice crosses. Crop Sci. 30:1178-1182.
Yeh, W.J. and Bonman, J.M. 1986. Assessment of partial resistance toPyricukria Oryzae in six
rice cultivars. Plant Pathol. 35:319-323.
Yu, X.P., Wu, G.R. and Hu, C. 1990. Studies on tolerance and anitbiosis nature of rice
varieties to whitebacked planthopper. Acta Phytophylactica Sínica 17:327-330.
Zeigler, R.S., Thomas, Nelson, R.J., Levy, M. and Correa Victoria, F.J. 1994. Lineage
exclusion: A proposal for linking blast population analysis to resistance breeding. In: Rice
Blast Disease, R.S. Zeigler, S.A. Leong, and P.S. Teng, (eds.). IRRI, Los Baños, Philippines.
Zhang, G., Angeles, E.R., Abenes, M.L.P., Khush, G.S. and Huang, N. 1996. RAPD and RFLP
mapping for the bacterial blight resistance genes Xa 13 in rice. Theor. Appl. Genet. 93:65-
70.

Breeding Rice for Resistance
to Diseases and Insect Pests
Ram C. Chaudhary*'
INTRODUCTION
Rice has been under cultivation for over thousands of years and in 115
countries. As a result, it has become a host for a number of diseases and
insect pests, 54 in temperate zone, and about 500 in tropical countries
(Swaminathan, 1979). Of the major diseases and pests, 45 are fungal,
10 bacterial, 15 viral (Ou, 1985), and 75 insect pests and nematodes.
DISEASES
The major fungal diseases of rice are blast, sheath blight, brown spot,
narrow brown leaf spot, sheath rot and leaf scald. Ten major bacterial
diseases have been identified in rice ({IRRI, 1969; Ling, 1972; Ou, 1985;
Goto, 1979, 1988). The bacterial diseases, which cause serious economic
losses in rice-growing countries are bacterial blight, bacterial leaf streak,
and bacterial sheath rot. Twelve viral diseases of rice have been
identified but the important ones are tungro, grassy stunt, ragged stunt,
orange leaf (in Asia), hoja blanca (America), and stripe and dwarf (in
temperate Asia).
* Ex~Global Co-ordinator INGER, International Rice Research Institute, Philippines;
Chairman, Participatory Rural Development Foundation, Gorakhpur 273014, India,

170 Rice Breeding and Genetics: Research Priorities and Challenges
Blast
HIM
'1.
Blast disease of rice caused by the fungus Magnaporthe grísea (anamorph^.
Pyricularia grísea) is the most destructive one. It has been reported from
almost all rice-growing countries of the world. The fungus can infect the
rice plant at any growth stage. Typical leaf lesions of leaf blast or foliar
blast are spindle-shaped. Large lesions (1.5 cm x 0.5 cm) usually develop
gray centers. The lesions collapse and leaves of the susceptible varieties
may be killed. Pinhead-size brown lesions, indicating resistant reaction,
may be confused with the symptoms of brown spot. The fungus may
also attack the stem at the nodes, node blast, in which the stem bends
and breaks at the nodes causing complete spikelet sterility. The fungus
may also attack the last internode, neck blast, causing partial to complete
sterility.
G enetics
Leaf blast
Chaudhary and Nayak (1987) and Gangopadhyay and Padmanabhan
(1987) have reviewed the inheritance of leaf blast. Goto (1978) observed
a decline in resistance in Ginga, a lowland variety and descendant of
Sensho. Sensho, an upland variety possessed a high degree of blast
resistance with Rbl, Rb2 and Rb3 genes, but Lazy Ginga, La-isogenic
line of Ginga, proved that Rbl of Sensho had not been introduced to
Ginga. Absence of Rbl caused a significant decline in blast resistance of
Ginga. The two blast resistance genes, which controlled a moderate level
of resistance in Ginga, were assumed to be multiple alleles or the two
genes other than Rbl of Sensho.
Kiyosawa and Cho (1980) made a detailed study of blast resistance
in Tongil. The cross of Palkweng/Tongil in F3 showed segregation,
which could be explained by the two or more genes against seven fungal
strains, which had been used as differential strains in Japan. The tests of
hybrids of Tongil with Kiyosawa^s differential varieties indicated that at
least two of these genes for resistance were Pi-a Pi-b. The test of Tongil
with fungus mutants.supported the presence of two genes.
Kiyosawa (1981) put forth the concept of gene-for-gene for blast
resistance in Palkweng-Tongil against Ina-72 and Ina-72+ races. The
possibility of application of the gene-for-gene concept to the hostpathogen
relationship of rice and rice blast system was confirmed
through gene analyses in Japanese rice varieties. According to the genefor-gene
concept in blast, resistance to virulent isolates of P. grísea finds a
susceptible genotype in the resistant cross but not in the resistant parent.
It was also found that when the parents are resistant to an isolate, their
descendants become susceptible to the same isolate (Goto, 1978). Ginga,

Ram C. Chaudhary 171
a descenant variety of Sensho and Fukunishiki^ parental cultivar Zenith,
was found susceptible to some races of P. grísea whereas their parents
showed resistance. The authors concluded that the resistance gene linked
with a marker could not be inherited, so the descendahts failed to retain
the resistance as in their parents.
Kiyosawa et al. (1983) developed mathematical models to study the
resistance gene frequencies in each prefecture of Japan. Resistance to
blast disease between a hybrid Fukunishiki and its parental cultivar
Zenith was found to be different (Goto, 1976). Both varieties carried the
major gene Pi-z but Fukunishiki had one recessive gene and Zenith had
the modifier Rb6 to the blast strain Ken-53-33, Rb6 , the modifier gene of
Pi-z in Zenith, gave effective control for isolates Fuku 67-57 and Ken-53-
33 while Fukunishiki, the derivative of Zenith could not possess Rb6 . It
was also formd to be less resistant than Zenith. Kiyosawa and Yabuki
(1976) observed the presence of Av-a, Av-k'*’, Av-a'^, Av-k and Av-a^,
Av-k genes, but not Av-a Av-k gene in Kanagawa Prefecture, Japan
where Pi-a+ Pi-k, Bi-a Pi-k+ and Pi-a Pi-k+ varieties were grown. The
authors developed a model on the race frequency change in a hostpopulation
system with genes for resistance and avirulence gene for
host. They assumed genotypes of true resistance, Ab, A+, +B, ++, in a
host population and four genotypes (races) for avirulence, ab, a+, +b
and tt in an airborne pathogen population. They established equations
following the rate of frequency changes on a given, host population. In
this case, A and B were resistance genes while a and b were avirulence
genes which specifically corresponded to the respective resistance genes.
Kiyosawa (1981) studied the race frequencies of pathogens or
virulence analysis in blast fungus by computer simulation and
theoretical equations. The ratios between the observed value to the
expected value of the fungus genot5rpe Av-i Av-k+ (abbreviated "tt"
genotype ratio) for virulence was calculated. The "tt" genotype ratios
were higher in different districts of Japan for all the ten years. The "tt"
genotype frequencies showed decreasing values with decrease of
individual virulence genes. The change in frequencies of the "tt"
genotype (race) could not be explained by simple directional or
stabilizing selections. The "tt" genotype ratio returned to one even after
directional and/or stabilizing selections and that limited use of virulence
(v) gene analysis to search for the causes of the change in pathogen
geirotype (race) frequencies.
Padmanabhan (1965) studied the inheritance of leaf blast resistance
in the cross Co. 13 Co. 25 and its reciprocal. They found the resistance in
Co. 25 to be controlled by three genes along with modifiers. Rath and
Padmanabhan (1972) reported the presence of one independent
dominant gene in Zenith and Tetep against race lA-II and ID-I and one

172 Rice Breeding and Genetics: Research Priorities and Challenges
dominant gene in Ginga and Norin to the race lA-II. Padmanabhan and
his colleagues (Padmanabhan^ 1974) did a detailed study using two
races, viz. IDl and IC17 of P, grísea, 16 crosses with IDl and 8 crosses
with IC17. The was resistant in crosses involving the resistant parents.
Zenith S 67, Tetep and Tadukan and susceptible parents, Co 13.
However, in the cross Cl 5309 (S) and S 67 (R), the Fj was resistant to
race IC 17. Fj was susceptible when tested against IDl. Against race IC
17, the cross involving Cl 5309 and Zenith revealed that the inhibitory
gene in Cl 5309 and Zenith was selective in action and expressed itself
only with some races of P, grísea, confirming the earlier findings of Rath
and Padmanabhan (1972). Studies with F2 and F3 showed that the
resistance of Tetep, Tadukan, Zenith to the races IC17 and IDl was
governed by three pairs of dominant genes of which any two could
confer resistance to the varieties. The three genes might be the same or
allelic in the varieties. The resistance of S 67 to race IC 17 and ID 1 was
governed by two pairs of dominant complementary genes which were
also present in Zenith but not expressed tmder predisposing conditions
of high fertility and low night-temperature. The variety Cl 5309
appeared to have four pairs of genes including one pair of inhibitory
genes which inhibited the resistance gene of Zenith, Tetep, and Tadukan
with respect to IC 17 of P. grísea. The variety BJl might possess general
or field resistance governed by polygenes. Based on these studies, the
genetic constitution of the varieties to race IC17 and ID 1 of P. grísea
were designated as: Zenith (Pi-za, Pi-zb, Pi-zc), Tetep (Pi-za, Pi-zb, Pizc),
Tadukan (Pi-za, Pi-zb, Pi-zc), S 67 (Pi-za, Pi-zb or Pi-zc), BJ 1 (Pi-za,
Pi-zb or Pi-zc), Cl 5309 (Pi-za, Pi-zb, Pi-zc (IP 1-z), Co. 13 (Pi-za, Pi-zb,
Pi-zc (IP 1 -z).
The genetic constituents of rice varieties for blast resistance has been
estimated by utilizing Japanese cultivars and Pi-k, Pi-a, and Pi-z and Pik,
Pi-ta and Pi-ta2 resistance genes in Tetep, Zenith, and Tadukan
respectively have been assumed. But Mohanty and Gangopadhyay
(1982) observed digenic blast resistance in Zenith Tadukan, while
monogenic resistance in Tetep to Cl isolated from F2 population to their
crosses with Ratna, Karuna and Co. 13. Tetep, Zenith and Tadukan
possess one, two, and three genes respectively for blast resistance to C3
isolate of P. grísea.
Methods for Gene Analysis
The Mendelian ratio method of gene analysis is followed in most
countries. In this method, usually one resistant variety belonging to a
particular group is crossed with a susceptible variety and the F2 hybrids
are artificially inoculated with a fungus strain. The pattern of resistant
and susceptible ratio showed the gene involved resistance. F3 seedlings
consisting of resistant and susceptible progenies were raised in rows

Ram C. Chaudhary 173
and tested against separate fungus strains. When all the plants in the F3
lines showed resistant reactions to one fungus strain^ it was concluded
that the same resistance gene controlled the resistance against the two
fungus strains. Different reactions of some lines to the two fungus
strains indicated participation of different genes in the process.
Kiyosawa (1976,1978) devised the frequency distribution curve method
in order to avoid the great deal of labour required in gene- analysis
through hybridization.
Quantitative Inheritance
Rath and Padmanabhan (1972) in their genetic analysis of the penetration
and establishment phases of the disease reaction concluded that during
the penetration phase field resistance was controlled by a polygene with
few loci. Using the sheath inoculation technique, Kaur et al. (1984)
reported that penetaration and an establishment phase was governed by
polygenes, while resistance to spread of infection in the host was
governed by major genes. Takahashi (1965) proposed a working
hypothesis to explain the relationship between gene action and
expression of resistance for the rice blast disease and indicated that (a)
there were several pairs of genes controlling blast resistance in rice
varieties^ (b) there were races of pathogen specific in their host reaction,
(c) fewer fungal races were associated with high degree of resistance, (d)
a susceptible host cell permitted symbiotic mycelial growth for a fairly
long period while a resistant cell inhibited it by a hypersensitive
reaction, and (e) mycelial growth was inhibited in the infected cells of
the host due to some factors resulting from the host-parasite interaction.
Quantitative analysis of the lesion type demonstrated distinct
resistance/susceptible reactions in the parents involved; the mean value
of the F2 progenies of the different crosses were closer to the parental
values rather than the corresponding mid-parental values. This
suggested that major genes might be responsible for this phase of disease
reaction. Quantitative estimations for the lesion number showed that the
variety Tetep was highly resistant, Zenith was resistant, wliile the other
five were susceptible to both the races imder study. Mean values of F2
populations in most cases were intermediate between the parental
values and a few were in the extra-parental range. This suggested that
the factors controlling lesion number were primarily polygenic.
Hashioka (1950) examined the degree of resistance on the basis of
lesion types. Takahasi (1965) steered the need for quantitative analysis
to estimate blast resistance and did a critical study on the basis of
number and type of lesions, independently and in combination and on
the measurement of the degree of mycelial growth in host cells. He
observed the method based on the degree of mycelial growth in the host
cells to be accurate in estimating resistance to blast.

174 Rice Breeding and Genetics: Research Priorities and Challenges
Linkages
■ !i ?í
ií.- ■■■ I
Hr
The early variety of Koshihikari and the late variety Shiranui were
crossed with the testers ER and LR which carried alleles of Lm locus for
earliness and lateness respectively, as well as Pi-z4 for resistance to P.
grísea (Yokoo and Kikuchi, 1977). Shiranui appeared to carry Lms, which
conferred earlier heading than "Lmti" and at another locus, a gene
conferring late maturity than "Lm4". The segregation ratio of early to
late in the F2 of Koshihikari Shiranui was 1:3 as expected from the
contributions of "Lme" and "Lms". Yokoo and Kikuchi (1977) also
reported the linkage relation between the heading time and blast
resistance of seven varieties with early maturing (ER) and a late
maturing lime (LR) carrying Pi-zt for resistance to P. grísea. "ER" carried
the "Em2" allele for heading time and the "LR" carried the "Lm ". The
heading time of early varieties was controlled by "Lm " alleles for
lateness. "Lm" was closely linked to Pi-zt. Goto (1976) also showed that
resistance to blast of Fukunishiki derived from Zenith was the same as
that of Zenith. Both varieties carried the major gene Pi-z but Fukunishiki
had one recessive gene and Zenith had a modifier, RB6 , to the blast
strain Ken 53-33. Rb6 , the modifier of Pi~z in Zenith, expressed resistance
to isolate Fuki 67-57 and Ken 53-33, while Fukunishiki without Rb6 was
less resistant than Zenith.
Goto (1978) observed that Ginga, derived from Sensho (an upland
variety) which had high blast resistance with three genes, Rbl, Rb2, Rb3,
did not have Rbl. Rbl gene of Sensho linked with gene la had not been
introduced in Ginga. Goto (1970) followed the sheath inoculation
method and found three independent genes, Rbl, Rb2, Rb3, which acted
additively for resistance to the four isolates belonging to three
international races, IC 1 (Ken 53-33), IG-1 (Hoku-h Hoku 2 ) and IF 1
(Nagasa) in a two-cross combination, Sensho, H 79. tujimaki et al. (1975)
observed eight indica varieties collected from four Asian countries to be
resistant to seven differential strains of P. grísea selected by Kiyosawa. A
single dominant gene, Pi-zt, controlled the resistance in Ken 53-33. This
gene was widely distributed among the indica varieties tested.
Kiyosawa (1972a) transferred the resistant gene from indica varieties to
B T 8 , Bl-9, and Bl-11. These varieties carried the independent gene Pi-b
and Pi-t. Either of these gene pairs were closely linked to Pi-a, Pi-k, Pi-i,
Pi-ta, Pi-z, Pi-m, or Pi~f.
Yokoo and Kikuchi (1977) made crosses among five varieties that
varied in heading time and presence of the Pi-zt allele for resistance P.
grísea. In the F2 of Fujisaki 5 J315, homozygous resistant plants were
mostly late heading, homozygous susceptible plants were early heading,
and heterozygous plants were mostly intermediate in heading time.
Linkage between heading time and resistance was observed in Norin 8
J315.

n
Ram C. Chaudhary 175
Goto et al. (1981) analyzed the linkage of four genes^ viz. Pi-a, Pi-k,
Pi“Z, and Pi-I and reported that Pi-a and Pi-k were in rectilinear order in
the 8 th linkage group and Pi-I and Pi-z were independent of the first
group. Studies on linkage relationship showed that all the genes for blast
resistance found so far belong to the four linkage groups^ "bt^^ "la"
and "wx". Prom the results obtained in a uniform blast nursery triab
Carreon and Tetep were identified as parents with a broad spectrum of
resistance but Carreon proved to be a poor combiner. Progeny of IR 1416 .
(14-400 Tetep) and Progeny of IR 1544 (IR 24 Tetep) have excellent grain
quality of good combining ability and are resistant to green leafhopper.
Many lines with blast resistance and multiple tolerance to insects and
diseases were developed. Gompai-15 was used in crosses and three
blast resistant lines with multiple tolerance to other diseases and insects
were developed and named as IR 28, 29 and 34.
According to Li et al. (1983), anther culture was found useful in
introducing Pi-zt gene from Toride 1, Toride 2 into Zhonghua 8 and
Zhonghua 9. This could be accomplished within a year while the
backcross method took 12 years. Colour "c" cluster spikelets "cl" and
white striped "ws" waxy endosperm "wx", brown mottled discoloration
of leaf and panicle gold hull "b ll", triangular hull "ki", Waise-aikoku,
Shirosasa dwarf "dw" (Iwata and Omura, 1971) were found to be linked
with resistance.
Kiyosawa (1972b) compiled data on the linkage relationship among
genes for blast resistance and other traits in four groups; taking into
consideration all the information available upto that time. Introduction
of useful genes by backcrossing was developed by Fujimaki et al. (1975).
When Pyricularia resistant, photosensitive, and late flowering indica
varieties were backcrossed as donor parents with susceptible japónica
varieties, resistance and days to heading were polygenic in Norin
25/Co. 4, Morak Sappillai/Fujisaka 5 and Fujisaka 5/Konotor. In BCj
generation of Norin Tjina and four other crosses, segregation for
resistance was digenic, while in BC2 it was monogenic or digenic.
Kiyosawa (1968) studied the genetics blast resistance in the Chinese
varieties, To-to type, Chokoto, and Minohakai and reported the presence
of genes, Pi-a and Pi-k. Michakasi was found to possess an additional
gene, Pi-m. No linkage relationship was found between Pi-a and Pi-k,
Pi-a and Pi-m, and Pi-m and Pi-k. In the Korean variety, Diazichall
(Butamochi), Kiyosawa (1968) found Pi-a and Pi-i genes after testing
against seven fungus strains. The gene Pi-i behaved independently of
the gene Pi-a.
Kiyosawa (1969) studied the genetics of resistance in two varieties,
Shiriniki and Kusabe varieties using the Philippines race Ken ph. 03 and
its mutant Ken ph. Od2 and reported the presence of two dominant

176 Rice Breeding and Genetics: Research Priorities and Challenges
genes for resistance, Pi-k2 and Pi-k, in Shiriniki and Kusabe, The gene
Pi-ks was allelic to the gene, Pi-k. Gene pattern for resistance in the
Pakistani variety Pusur was complex and conditioned by three genes
(Kiyosawa, 1968). Of the three genes in the Var. K2 (derivative of Pusur),
two were identical and or allelic to Pi-k and Pi-s and were designated Pik6
. Thus three genes (Pi-k, Pi-ks and Pi-kl) of the nine known genes of
Japan were located on one locus.
Analysis of hybrid progenies of HR 22 by Kiyosawa and Murty
(1969) indicated that a gene similar to Pi-k controlled resistance in five of
the seven strains of P. grísea tested. A derivative from the hybrid HR 22/
Sasashigure was analyzed. The gene for blast resistance was observed to
be an allele of Pi-k and was designated Pi-kh.
Nagai et aL (1973) were the first to introduce the variety Toride 1
(Toride is an indica and japónica cross with resistance gene from TKm 1)
of India as a multiracial resistant donor parent. After backcrossing four
times with Norin 8 of Japan as a recurrent parent, they succeeded in
producing the variety TKM 1. This selection proved resistant to most of
the major pathogenic races of the rice blast fungus collected throughout
the country, Yokoo and Kiyosawa (1970) reported that the resistance in
TKMl was controlled by a dominant gene, designated Pi-zt, which
differed from Pi-k, Pi-i, Pi-ta, Pi-a and their alleles. The resistance gene
of Toride 1 was allelic to Pi-z of the USA variety Zenith and it seemed to
be at a different locus from that of Pi-m.
Blast resistance of Toride 2, a variety with multiple resistance, bred
from the hybrid of (Norin 8 Co. 25) and (Norin 8 ) was analyzed
(Kiyosawa and Yokoo, 1970). Toride 2 carried the genes Pi-zt and Pi-a.
The gene Pi-zt was the same as that found in the variety Toride 1, which
was bred by transferring the resistance gene in the Indian variety TKM 1
to Norin 8 . Following the sheath inoculation technique. Goto (1970)
carried out gene analysis for blast resistance in two cross combinations,
viz. Sensho H79 and Imochi Shrirazu H9 with four isolates belonging to
three international races, IC-1 (Ken 53-33), IG-1 (Hoku 1 and Hoku 2 )
and IF-1 (Naga 8 a). Three independent genes, Rbl, Rb2 and Rb3 acted
additively for conferring resistance to the four isolates.
Nagai et ah (1973) described in detail the blast resistance of variety
Toride 1 and Toride 2 to seven strains of P. grísea. Toride 1 was found to
carry a new resistance gene Pi-z+ that was independent of Pi-i, Pi-k, and
Pi-ta, but weakly associated with Pi-a and allelic or closely associated
with Pi-z-I-. Resistance in Toride 2 was controlled by Pi-a and Pi-z+.
Kiyosawa (1970) reviewed the resistance of local and exotic varieties for
genetics of resistance of P. grísea. The resistance gene Pi-a was
universally distributed in the resistant varieties throughout the world.
The locus Pi-k had multiple alleles, Pi-k, Pi-k8 , Pi-kp, and Pi-kh, of

Ram C. Chaudhary 177
which only Pi-k8 was found in Japanese varieties. Transfer of resistance
genes to some indica varieties was carried out by Kiyosawa (1972a). It
was shown that the rice lines BL 8 / BL 9 and BL 11 carried the gene Pi~b
for resistance to P. grísea and BL carried the independent genes Pi-b and
Pi-t. Neither of these genes appeared to be closely linked to Pi-S/ Pi-k, Pii,
Pi-taj, Pi-z+/ Pi-m, or Pi-f. The presence of halo lesions appeared to be
controlled by Pi-t.
Kiyosawa (1974b) assessed 12 varieties for their response to seven
strains of P. grísea and classified their response pattern as Shin 2 (New 2)
and Shinsetu or Sekhishunakumo types. In Shin 1 type^ a dominant gene
named Pi-i was found while Shinsetu type possessed Pi-i and Pi-a genes.
Kiyosawa (1974c) summarized all the possible genes in Japanese
varieties of rice against seven strains of P. grísea and divided them into
twelve groups. Some non-Japanese varieties were analyzed for their
reaction to P. grísea by crossing them with Japanese varieties and 13
resistance genes were identified in both Japanese and non-Japanese
varieties^ viz. Pi-a^ Pi-i^ Pi-ks^ Pi-kp, Pi-kh, Pi-ta^ Pi-ta2/ Pi-z, Pi-z+, Pi-m^,
Pi-b and Pi-t. The loci Pi-k and Pi-m were linked and so were Pi-z and Pii.
Kiyosawa (1983) suggested frequencies of nonrandom association of
virulent genes Av-k+ and Av-km and Av-ta+ and Av-ta2+.
This clearly indicates that not only matching of resistance gene, but
also an additive part should be present on the same loci for durable
resistance.
Diallel analysis
Diallel analysis has been employed by Mohanty and Gangopadhyay
(1985) to explore the nature of gene action for blast resistance in Tetep,
Zenith, Tadukan, Ratna, Jaya, Ratna, Karuna, and Co 13. Resistance in
the first three varieties was associated with the dominance effect and in
the last four with the recessive effect. The parental order of strength of
functional blast resistance genes was Zenith Tetep, Tadukan, Jaya,
Ratna, Karuna Co 13. Rath and Padmanabhan (1973) reported that the
lesion types were controlled by major genes and lesion number was
under polygenic control.
Wu et al. (1981) studied the genetics of resistance of Ta-poo-Choo-2,
Tadukan, and Mamoriaka, which were resistant to four races in Taiwan,
by crossing each of the four varieties with Lomello with no known gene
for resistance. Atkins and Johnston (1965) studied the crosses between
varieties Zenith and Gulfrose susceptible to race 1 and resistant to race
6 . They concluded that resistance to US race 1 and 6 was controlled by
an independent single dominant gene, Pi-1 in Zenith and Pi- 6 in
Gulfrose. Currently used donors for resistance, identified genes, and
their chromosomal locations are listed in Table 9.1, while improved
varieties released using various resistance genes are listed in Table 9.2.
i I

178 Rice Breeding and Genetics: Research Priorities and Challenges ■
Neck blast
Hashioka (1950) observed that resistance to neck blast was dominant in
some crosses while recessive in others. He was of the opinion that leaf
blast resistance was independent of neck blast. Ou and Nuque (1963)
artificially infected several varieties in the leaf and half-emerged panicle
stages. They found that varieties resistant to isolates at the seedling
stage showed no neck blast while those susceptible to isolates at the
seedling stage had 46-100% neck rot.
Table 9.1
Rice genes conferring resistance to blast fungus, registered with the Rice
Genetics Cooperative {Rice Genetics NexiKletter, vol. 12,1995)
|i|' ^^
Resistance
gene
Donor variety
'Chromosome
location
Reference
Pi a Aichi Asahi 11 Yamasaki and Kiyosawa, 1966
Pi-I Ishikari Shiroke, Pujisaka 5 6 Yamasaki and Kiyosawa, 1966
Pik Kanto 51, Kusabue 11 Yamasaki and Kiyosawa, 1966
Pi k-m Tsuyuake 11 Kiyosawa, 1968
Pi k-p Pusar, K60 11 Kiyosawa, 1969
Pi k-h HR-22, K3 11 Kiyosawa and Murty, 1969
Pi k^s Shin 2, Norin6 11 Kiyosawa, 1969
Pi z Zenith, Fukunishiki 6 Kiyosawa, 1967
Piz-t TKM l,Toridel 6 Yokoo and Kiyosawa, 1970
Pi-zS*^ 5173, C101A51 6 Mackill and Bonman, 1992
Pi ta Tadukan, Pi No. 1, K 1, 1 2 Kiyosawa, 1966
Pi ta-2 Tadukan, Pi No. 4, Reiho 12 Kiyosawa, 1967
Pib Bengawan, BLl 2 Kiyosawa, 1972
P it Tjahaja, K59 1 Kiyosawa, 1972
Pish Shin2, Nipponbare - Imbe and Matsumoto, 1985
Pi-1 LAC23, CIOILAC 11 Mackill and Bonman, 1992
Pi-2 5173 6 Inukai et«/., 1994
Pi-3 Pai-kan-tao, C104PKT 6 ? Mackill and Bonman, 1992
Pi-5(t) Moroberekan 4 Wang et ah, 1994
Pi-6 (t)* Apura 1 2 Yu et al, 1991
Pi-7(t). Moroberekan 11 Wang et aL, 1991
Pi-8 Kasalath 6 Pan et ah, 1996
Pi-9(t)* WHD-IS-75-1-127,0 . minuta - Reimers et at., 1997
Pi-lO(t)* Tongil 5 Naqvi and Chattoo, 1996
Pi-ll(t)* Zaiyeqing 8 8 Kinoshita et ah, 1996
Pi-12(t) Moroberekan Inukai et ah, 1996
Pi-13(t) Maowangu 6 Vaxietah 1995
Pi-14(t) Maowangu 2 Pan et al., 1995
Pi-15(t) GA25 - Pan et ah, 1995
Pi-16(t) Aus373 2 Anon., 1995
Pi-17{t) DJ123 7 Anon., 1995
Not yet registered
In the Uniform Blast Nursery in Taiwan, Chang et ah (1965) reported
that neck rot and seedling blast were fbund to be different and perhaps

m i
Ram C. Chaudhary 179
separate genes acted for resistance in leaf and neck blast resistance. In
the cross between Co. 13 and Co. 25 and its reciprocal cross, neck blast
resistance was observed to be governed by two or three dominant genes
with modifiers (Padmanabhan, 1965), who suggested that genes for leaf
blast and neck blast resistance were different. Ito (1965) could find no
correlation between neck and seedling blast resistance. Ou (1985)
emphasized the existence of different races for which plants showing
leaf blast resistance later became susceptible to neck blast. In the field,
many races might be present and they might change at different seasons.
When varieties are thoroughly tested for resistance at the seedling stage,
further testing at the flowering stage does not appear necessary.
Table 9.2
Summary of resistance and susceptibility of IR varieties developed by
IRRI, Philippines.
Variety
Blast
Bacterial
blight
Disease and insect reactions
Tungro
Green ■
Leafhopper
Brown
Planthopper
Stem
borer
Gall
midge
IR5 MR S S R S MR S
IRS S s S R S S S
IR20 MR R s R s MR s
1R22 S R s S s S s
IR24 S S s R s S s
IR26 MR R MR R R MR s
IR28 R R R R R MR s
IR32 MR R R R R MR R
IR36 MR R R R R MR R
1R38 MR R R R R MR R
IR42 MR R R R R MR R
1R46 MR R R R R MR R
IR50 S R R R R S -
IR54 MR R R R R MR -
IR58 MR R R R R S
„
IR60 MR R R R R MR -
IR62 MR R R R R MR -
IR64 MR R R R R MR -
IR66 MR R R R R MR -
IR68 MR R R R R MR
„
IR72 MR R R R R MR -
R: Resistant; S: Susceptible; MR: Moderately resistant.
In Egypt, Balal et al. (1977) studied leaf blast and neck blast in five
crosses from varieties resistant (YNA 282 and Arabi) and susceptible
(Nahda and Giza 159), Observations in the F2 and F 3 indicated that leaf
blast resistance was simply inherited and controlled by two genes
designated LPl and LP2. YNA 282 possessed LPi while Arabi carried
Lpi Ppi; Nahda and Giza 159 both possessed IPil and IPi2 genes.

180 Rice Breeding and Genetics: Research Priorities and Challenges
Resistance to leaf blast was positively correlated with neck blast in three
out of four crosses. The correlation was highly significant. YNA. 282
possessed the complementary genes NPil and NPi2 while Arabi carried
NPi3. Resistance to leaf blast was highly and significantly correlated
with resistance to neck blast in three out of four crosses in a positive
way. Hence^ the genes for leaf blast resistance (LPil and LPi2) were
effective to some extent in combating neck blast.
B r e e d in g
iiíi' :
Breeding for blast resistance has been followed in various countries for
the last 60 years. Efforts and achievements of early years have been
reviewed by Ito (1965)/ Padmanabhan (1965)/ and Chaudhary and Nayak
(1987). Broadly speaking, there are two types of resistance: avoidance or
disease escape and true resistance. Avoidance involves a heterogeneous
group of mechanisms, which depends on the structure of the entire
plant/ of certain plant parts or of certain plant tissues. True resistance
reduces the first moment of contact between the host tissue and the
parasite. It implies an intimate contact between the host tissue and the
parasite and involves the expression of their mutual interaction.
Resistance may be of a hypersensitive type, which is frequently used in
resistance breeding. The resistance controlled by major genes is race
specific and has been used extensively. But due to hitherto unexplained
reasons, such resistance has had a short life span. Contrary to major
genic resistance, the partial resistance, controlled by polygenes, is
durable. Thus in the breeding program, breeding lines or test varieties
with less than complete resistance are selected.
Over time, various strategies have been followed to control blast
disease using host plant resistance. The usual practice of replacing the
succumbed variety has been followed but with little success. Pyramiding
of the major blast resistance gene was followed but given up for fear of
developing the super race of "blast fungus".
Ezuka (1979) suggested that if a high level of "field resistance" were
to be incorporated in a variety with "true resistance", such a variety
would be more stable. Resistance genes Pi-ta and Pi-ta2 from the
Philippines variety Tadukan were transferred to japónica background.
A number of varieties resistant to blast were released in Japan: Shinju,
Futaba, Vasevakaba, Koganennishiki, Ukonnishiki, Hoiriarenishiki,
Fugimoriri, Reimi, Kanto 51, Kanto 53, Kanto 54, Kasabue, Yuukara,
Shenshuraku, Kongo, Minehikari, Shikokata, Tosa, Senbon, Asahikari,
Satomino, Akige, Toride 1, Toride 2.
In India, Co. 4 was developed as a resistant variety in 1924 followed
by TKM-1. Since blast disease is limited to hilly tracts and certain

Ram C. Chaudhary 181
seasons, breeding is limited to just those situations. From time to time a
number of varieties have been released for such situations. At IRRI,
Philippines, incorporation of blast resistance is one of the important
aims of the breeding program. Donors such as Dawn, Tetep, Zenith,
Gam pai 15, Pankhari 203, Carreon, Ram Tulasi, Oryza nivara, and a
number of improved plant type lines are used on a regular basis.
Screening of all breeding lines is a regular feature. As a result, most test
entries from the breeding program come out with a reasonable degree of
success. In addition to the concept of complete resistance, partial
resistance is followed in the irrigated rice ecosystem with the idea that
such resistance may be durable,
Given the uncertainty of the variability of the pathogen and the
history of resistance breakdown, it is not surprising that a number of
different plant breeding approaches have been proposed to achieve
durable blast resistance. Until very recently, however, the tools to apply
the concepts for breeding for durable blast resistance have been lacking.
Molecular techniques for characterizing pathogen virulence, pathogen
populations, and host plant resistance, should offer the means to test
hypotheses on the nature of durable resistance.
Proposals for the use of major genes have moved beyond developing
lines with single genes effective against a few pathotypes. Robinson's
(1973) and Nelson's (1978) general proposal that major genes may be
combined, or "pyramided", to confer "horizontal" resistance effective
against all pathotypes of a pathogen, was supported by Ou (1979) as a
means to achieve reasonably durable resistance to blast. Crill et aL (1981)
proposed that varieties with different major genes could be rotated in
and out of production in a particular region to take advantage of
virulence shifts in the pathogen population driven by the changes in
varietal resistance. Mackenzie (1979) proposed the use of multi lines, or
a mixture of lines, each carrying one or a few major genes. He
emphasized the care with which the resistance genes should be
manipulated, and that usefulness and durability will depend on how
they are deployed.
Other rice scientists have argued that major genes obscure the
underlying resistance of the plant and that the use of these genes may
result in lines that are extremely susceptible when the major genes are
overcome (the so-called "vertifolia effect" of Van der Plank). That is,
there will be no means to select for the minor genes, and in crosses in
which the parents differ markedly in composition there is a high
probability that most minor genes of interest will be lost. We can
presume that this underlying resistance is functionally equivalent to that
mediated by QTLs. Various proposals have been put forth to avoid
erosion of resistance, escape, and the "vertifolia effect". These include

182 Ríce Breeding and Genetics; Research Priorities and Challenges
M:
breeding in environments that are extremely conducive to blast disease
development ("hot spots"), rejection of breeding lines that are
symptomless, inclusion of traditional and "unimproved" sources of
resistance obtained from blast conducive environments, and proper
characterization of the target environment (Kiyosawa, 1972b; Ezuka,
1979; Ou, 1979; Buddenhagen, 1983; Bonman et al., 1992, Johnson and
Bonman, 1993). An approach integrating most of these suggestions has
been applied with some success at the Centro Internacional de
Agricultura Tropical (CIAT) rice breeding research farm in eastern
Colombia (Correa-Victoria et at., 1992) where materials are developed
for distribution to rice programs in Latin America, Latin America tends
to have serious blast problems because nitrogen application rates are
high, water control in irrigated fields is usually poor, and vast areas are
grown under upland conditions. A word of caution should be raised
regarding the use of "hot spot" breeding sites. By developing complex
resistance under conditions of high pathogen diversity and unknown
variability,'breeders may be allowing the pathogens to adapt to novel
resistance gene combination.
In contrast to CIAT, IRRI has opted to focus on developing partial
resistance for the less blast-conducive irrigated rice environments.
Breeding lines are evaluated in blast nurseries (Bonman et al., 1991,
1992) and those with relatively reduced disease development are
selected as partially resistant. Cultivars developed in this manner have
generally performed well in the field, but when conditions are
particularly favorable for disease development, economic losses may
occur.
Bacterial Blight
Bacterial blight (BB) caused by the bacteria Xatithomonas campestris pv.
oryzae. The disease has been reported from most Asian and African ricegrowing
countries. The disease is reported to produce three types of
symptoms: leaf blights, kresek, and pale yellow (Ou, 1985; Mew and
Vera Cruz 1977, 1979; Mew et ah, 1971). The leaf blight symptoms are
characterized by small water-soaked spots or stripes or lesions at the
margin of the leaf blades, from the early tillering stage to the flowering
stage. The spots enlarge and form a wavy margin. The lesions can cover
the entire leaf blade and may even advance into the leaf sheath. In
severely diseased fields, grains may also be infected, lesions appearing
as discolored water-soaked spots on the glumes (Ou, 1985).
Kresek symptoms are characterized by withering of the leaves of the
entire plant at the early vegetative stage (2~4 weeks after transplanting).
Affected plants show marked stunting and soft roots that later may

Ram C. Chaudhary 183
become detached and float on the surface of the water (Goto, 1970).
Kresek is usually caused by bacterial invasion of roots during
transplanting or through cut ends of leaves. Pale yellow symptoms are
characterized by the production of pale yellow leaves, which are
normally green, and the youngest leaf is uniformly pale yellow to
whitish. Broad greenish-yellow stripes may appear on the leaf blade
(Ou, 1985). Pale yellowing is normally associated with early infection in
which the growing point remains alive but translocation systems are
blocked by bacterial mass in the xylem vessels, The pathological
relationships between the three symptoms of bacterial blight syndrome
are not fully understood. Although caused by the same pathogen, kresek
and leaf blight appear to be distinct and independent of each other. Pale
yellowing is a secondary effect of either leaf blight or kresek (Hsieh and
Chang, 1977; Mew et a/., 1971).
E t io l o g y
Farmers of Fukuoka area in Japan first noticed this malady in 1884
(Tagami and Mizukami, 1962) and thought it to be a physiological
disorder resulting from high soil acidity, as the oozing from the leaves
had an acidic reaction. In 1908, Takaishi found bacterial masses in the
dew drops, isolated the bacteria and successfully inoculated the leaves
but did not name the organism (Ou, 1985). Subsequent workers
identified the organism as Bacillus oryzae Hora and Bokura, Pseudomonas
oryzae Uyeda and Ishiyama, Bacterium oryzae (Uyeda and Ishiyama,
Nakatsi), Xanthomonas oryzae (Uyeda and Ishiyama), and very recently
Xanthomonas campestris (Dye et ah, 1980).The disease is now widely
distributed in almost the entire Asian continent (Mew and Khush, 1981),
Australia (Buddenhagen and Reddy, 1972), Latin America (Lozano,
1977), and in several African countries (Mew and Khush, 1981). Yield
losses ranging from 10% to 80% have been reported in Japan, Indonesia
(Reitsma and Schure, 1950), Ir\dia (AICRIP 1971; Rao and Kauffman,
1977), and the Philippines (Reys et al, 1982). The disease not only causes
yield reduction, but also lowering of grain quality. The extent of damage
from the disease depends on factors such as temperature, relative
humidity, rainfall, wind, spacing, and growth stage of the crop when the
disease occurs.
V a r ia t io n in P a t h o g e n ic it y
Japanese scientists in 1957 suspected the pathogenic variability of the
pathogen when a resistant variety Asakaze was severely infected by the
blight (Khush, 1977a). Based on the differential reaction of the varieties
in different locations in various countries, many reports appeared

184 Rice Breeding and Genetics: Research Priorities and Challenges
(Buddenhagen and Reddy, 1972; Ezuka and Horino, 1974;
Gangopadhyay and Padmanabhan, 1987). Several reports supported the
existence of pathogenic races of the bacterium (Gangopadhyay and
Padmanabhan, 1987; Mew and Vera Cruz, 1977, 1979). Currently strains
of X, campestris cv. oryzae are classified into four groups based on their
pathogenicity. Some countries have specific pathotypes .that are very
different from others.
S o u r c e s o p r e s is t a n c e
yni
Rice germplasm collections have been evaluated for resistance to
bacterial blight in several countries, notably Bangladesh, China, India,
Indonesia, Japan, Laos, Malaysia, Nepal, Philippines, Sri Lanka,
Thailand, Vietnam, and a large number of donors have been identified.
The resistant doors identified at IRRI Philippines originated in three
geographic regions. A large number came from northeast India,
Bangladesh, Nepal, which was designated as Gene Center 1, for bacterial
blight by Khush (1977a, 1977b). The other group came from Gene Center
2 , which consists of southern India and Sri Lanka. The third Gene Center
is Java and adjoining islands of Indonesia. Only a few donors came from
China, Laos, Malaysia, the Philippines, and Thailand.
A very novel source of resistance was identified in an accession of
Oryza longistaminata (Devadath, 1983), who claimed it was "immune" to
the races of the pathogen known till then.
G e n e t ic s
Inheritance of resistance to bacterial blight disease in rice has been
studied extensively in Japan (Skaguchi et ah, 1968) and IRRI (Murty and
Khush, 1972; Olufowote et ah, 1977; Singh et ah, 1983). The studies in
Japan indicated that the resistance of the Kidama group to isolates of
group I of the bacterium were controlled by a single dominant gene,
Xa-1. The virulence of the Rantaj-emas group had two genes for
resistance, Xa-1 and Xa-2. The gene Xa-2 conveys resistance to the
bacterial isolates of group II. Xa-1 and Xa-2 are linked with a
recombination value of 2.16% and are located on chromosome 11. No
variety with Xa- 2 alone has been identified although segregates with
Xa- 2 Xa - 2 genotypes have been obtained in the segregating populations
of crosses between varieties of the Rantaj-emas group and Kinmaze
group (Khush, 1977a).
Ogawa et ah (1978) found that the resistance of the Kogyoku group
and Java group of varieties to bacterial group V was governed by a
single dominant gene designated as Xa-kg or a gene very closely linked
with it. Xa-kg was inherited independently of Xa-w but it was closely

Ram C. Chaudhary 185
linked with Xa-1 with a recombination value of 2%. The resistance in
Jaya 14 was reported to be controlled by three dominant geneS/ Xa-l, Xa-
2, and Xa-kg. Yamada (1984) studied the genetics of resistance in IR28,
which was resistant to all five pathotypes of the bacterium present in
Japan. He reported that one major gene controlled the resistance in this
variety to bacterial group I and another major gene controlled its
resistance to bacterial group V, and the two genes were linked closely
with a crossover value of about 4%. Resistance to other pathotypes (1/ II/
IV) was controlled by minor genes or polygenes.
Studies at IRRI on the genetics of bacterial blight resistance have
shown that bacterial blight resistance in rice may be dominant or
recessive depending on the donor parent. Resistance in Sigadis is
governed by a single dominant gene. Resistance genes in Sigadis and
'fKM 6 are allelic but the gene in Bjl is different. Likewise/ Zenith and
B859A4-18-1 have the same gene for resistance while the gene in Wase
Aikoku 3 is non allelic. Varieties IR20 and IR22 and the breeding line
IR1529-680-3 possess the same dominant gene for resistance designated
as Xa-4 (Petpisit et al., 1977). IR 1330-3-2 and PelitaI/1 possess Xa-4/ and
Kale and CB II have xa-5 gene for resistance (Olufowote et al., 1977).
The dominant gene in variety Semora Mangga from Indonesia was
located at the Xa-4 locus but its expression was different from that in
other varieties possessing Xa-4 gene. The gene in Semora Mangga
imparted resistance only at the booting and post-flowering stages and
was designated Xa-4b/ in contrast to the Xa-4a present in IR22, which
conveyed resistance at all growth stages (Librojo et al., 1976). A
dominant gene conveying resistance to bacterial leaf blight at booting
and postponing stages showing a phenomenon of dominance reversal
was reported by Sidhu and Khush (1978) and named Xa-6.
Among resistant varieties/ gene Xa-4b was most widely distributed/
followed by xa-5 and Xa-4a (Sidhu et ah, 1979). Cultivars DZ 78 and
PI231129 possess Xa-7 and xa-8/ respectively. Xa-7 conveys resistance at
the booting and postbooting stages whereas xa-8 is effective at all stages
(Sidhu et d., 1979). Variety Sateng from Laos possesses another recessive
genC/ xa-9/ which conveys resistance at all growth stages (Singh et al.,
1983). Variety CAS 209 from Senegal possesses a new gene/ Xa-10/ which
conveys resistance to Philippine isolates and belongs to group II (Yoshimura
et al, 1983).
The recessive gene xa-13 located on chromosome 5 was identified
from variety BJl. Xa-14 was identified from Taichimg native 1, Xa-16
from Tetep/ Xa-17 from Asaminori, Xa-18 from Toyonishiki/ xa-19 from
XM 5/ xa-20 from XM 6, and Xa-21 from Oryza longistaminata (Anon
1995).

186 Rice Breeding and Genetics: Research Priorities and Challenges
B reeding
Breeding work for resistance to bacterial blight has been carried out in
Japan for 60 years and at IRRI for 30 years. National rice improvement
programs of Bangladesh, China, India, Indonesia, Malaysia, Nepal, Sri
Lanka, Thailand, Vietnam are now endeavoring to incorporate resistance
to bacterial blight into the improved varieties (Chaudhary and Nayak,
1987). The first bacterial blight resistant variety, Norin 27, was bred at
Kumamoto Breeding Center and released in 1946 (Mizukami, 1966).
Askaze was bred in the early 1950s and Hayatomoto in the early 1960s at
Kyushu Agricultural Experiment Station. Since then, several varieties
resistant to bacterial blight have been developed but all of them have the
Xa-1 gene for resistance (Toriyama, 1975). The breeding program for
incorporating resistance to bacterial blight was started at IRRI in 1964
and several resistant varieties have been released by the IRRI and the
Philippine Seed Board: IR20, IR22, IR26, R28, IR29, IR30, IR32, and IR34.
Varieties IR36, IR38, IR40, IR42, IR43, IR44, IR45, IR46, IR48, IR50, IR52,
IR54, IR56, IR58, and IR60 released by the Philippine Seed Board are also
resistant. All these IR varieties have Xa-4 for resistance to bacterial
blight, although improved breeding lines with Xa-5, Xa-6 , and Xa-7 have
also been developed. Efforts to develop varieties resistant to BB at IRRI
and in other countries have been summarized by Khush (1977a, b; and
1989).
Brown Spot
The brown spot disease caused by Cochliobolus miyabeanus (Ito et
Kuribayashi) Drechsler ex Dastur, is more commonly known by its other
scientific name, Helminthosporium oryzae Breda de Haan. It has been
known for the past 80 years and has been reported in all rice-growing
countries in Asia, USA and Africa (Ou, 1985). The fungus attacks rice
plants in all growth stages. The symptoms of brown spot disease (Ou,
1985) are most conspicuous on the leaves and the glumes. They may also
appear on the coleoptile, the leaf sheath, panicle and more rarely on
roots of young seedlings. Typical spots on the leaves are oval, about the
shape and size of a sesame seed, brownish with a gray or whitish center
when fully developed. They are relatively uniform and evenly
distributed over the leaf surface. The most devastating rice-yield losses
that led to the Great Bengal famine of 1943 in which a few million people
died of starvation, is attributed to the brown spot disease (Padmanabhan,
1973).

Ram C. Chaudhary 187
Etiology
The fungal spores after falling on the leaf surface germinate and
penetrate the tissue. Nawaz and Kausar (1962), Misra and Chatter] ee
(1963) noted that the pathogenicity of fungus isolates ranges from very
weak to extremely virulent.
Genetics
Emphasis centered on identifying varieties resistant to brown spot after
the 1943 epidemic 1943 in Bengal. A number of varieties were identified
as resistant in India^ viz. Dakar Nagar 273-32, Patnai 549-33, Kalma 219,
and Nagra 4 M 4 (Ganguli, 1946), Ch 13, Ch 41, Ch 45, T498-2a, Co 20,
BAM 10, T998M, T2112, T2118, and T96 (Padmanabhan et al, 1966). In
Japan, Asada et aL (1954) identified Hainan No. 217, and Chin Tiu Chin
as resistant.
The genetics of resistance to brown spot has not been studied
adequately. Nagai and Hara (1930) reported that resistance in a Korean
strain of rice was monogenic and dominant. On the other hand, Adair
(1941) found resistance to be controlled by a recessive gene. Studies at
IRRI, Philippines (IRRI, 1983), reported resistance to be controlled by a
single dominant gene in one variety, and by two complementary genes
and one inhibitory gene in another variety.
Breeding
Although no organized effort is currently employed, nonetheless
breeding lines are scored and selected for resistance under filed
conditions in most countries where brown spot is a problem. In India, all
the trial entries are screened on a regular basis (Misra et al, 1976) by the
Central Rice Research Institute (CRRI), Cuttack, and Directorate of Rice
Research (DRR), Hyderabad.
Sheath Blight
Sheath blight (ShB) of rice caused by Rhizoctonia solani Kuhn, once a
minor disease, has become a major disease in many countries, inflicting
heavy losses. It was first reported from Japan in 1910 by Miyake and
subsequently in Bangladesh, China, India, Indonesia, Philippines, Sri
Lanka, Thailand, USA, Brazil, Surinam, Madagascar, Malaysia,
Vietnam, and Venezuela (Premlatha Dath, 1990).
Sheath blight usually attacks rice plants at the tillering stage, causing
greenish-gray ellipsoid or ovoid spots (about 1 0 mm long) on the leaf
sheath. Sclerotia are formed on or around these spots but are easily
detached. The size and color of the spots and the formation of sclerotia

188 Rice Breeding and Genetics: Research Priorities and Challenges
depends on environmental conditions. Under favorable conditiorts they
are also formed on the upper leaf sheath and the leaf blades. Eventually
the whole leaf blade gets blighted^, killing most leaves partially or fully.
Grain formation and filling is severely affected. Singh and Pavgi (1969)
reported details of infection^ etiology and losses due to sheath blight.
Mizuta (1956) reported a yield loss of 25% when the blight affects the
flag leaf. Earlier the disease was serious in the temperate regions where
dew deposition was heavy for prolonged periods. But now due to the
use of high tillering varieties, high plant population and heavy use of
nitrogen fertilizers, the diseases has spread to other areas as well.
V ariation in Pathogenicity
Fungal isolates differ in pathogenicity (Akai et ah, 1960; IRRI, 1974).
However, the susceptible and resistant reactions used in distinguishing
the races are not so clear cut, often creating confusion.
Resistant donors
Among cultivars grown in the southern United States, the highest level
of resistance is found in the short and medium grain type rices, which
are closer to japónica types. The long grain types are mostly susceptible
(Lee and Rush, 1983). None of the rice varieties grown in Taiwan are
resistant. Since 1970, thousands of breeding lines and varieties have
been tested at IRRI, Philippines, using the adult-stage inoculation
method (IRRI, 1974) but none were found to be resistant (IRRI, 1984;
Premlatha Dath, 1990). Premlatha Dath (1990) has cited a large list of
resistant to moderately resistant lines; however, their performance was
not consistent under close scrutiny. This was primarily due to the
methods of inoculation, field conditions, inoculum used, disease scoring
method, aggressiveness of the pathogen used, etc. Resistance was also
reported in some accessions for O. barthii and O. rufipogan (Kannaiyan
and Prasad, 1978).
A number of factors such as morphological (tightness of leaf sheath
around stem, amount of wax deposition on outer leaf sheath) and
presence of toxins are associated with incident of the disease.
Inheritance of resistance
Limited information is available on the inheritance of sheath blight
resistance primarily due to a dependable resistant parent. But
indications from the performance of F^s are that the resistance is
dominant (Premlatha Dath, 1990). In certain cases a clear-cut 3 : 1
resistant : susceptible segregation pattern was reported (Hashioka, 1951;
Chang, 1962) in the F2 generation. Wax thickness, which was correlated

Ram C. Chaudhary 189
with resistance, also segregated into a 3:1 ratio in p2 - Goita (1985)
studied the inheritance of resistance in crosses between four susceptible
and three resistant lines. Inoculation was done under control conditions.
The frequency distribution of the F2 progeny showed a bimodal pattern
with modes at 5 and 7. In most crosses, the resistant-to-susceptible ratio
was 9 ; 7, suggesting that two pairs of complementary genes controlled
the resistance. Heritability estimates were low though, possibly due to
epistatic interactions and other unexplained reasons.
B r e e d in g
Due to the fact that dependable resistant donors are lacking, breeding
for resistant cultivars has not been very successful. Even in Japan where
the disease is serious, successful breeding for resistance has not been
achieved (Toriyama, 1975). Khush (1977b) suggested a breeding strategy
which involves continuous evaluation of the germplasm to identify lines
with higher levels of resistance, through screening of all advanced
breeding lines and elimination of highly susceptible material, and
organized breeding programs using selected lines with moderate levels
of resistance. The program aims at pyramiding minor genes for resistance
from several parents into the same line by recurrent selection.
But such approaches failed to yield expected results at IRRI
Philippines and Indonesia. A program to breed high yielding cultivars
with a high level of resistance was initiated in 1979 at DRR, Hyderabad,
using such donors as Pankaj, Ramadja, T 141, OS 4. Some breeding lines
were reported to combine better levels of resistance with high yield
(Rani and Satyanarayana, 1982; Reddy et al, 1986). The approach using
genetic engineering with chitinase genes from rice and barley appears
bright (Khush, 1998) for the control of ShB in rice.
Tungro
Tungro virus disease of rice is limiting rice production in Bangladesh,
India, Indonesia, Malaysia, the Philippines, and Thailand, causing yield
losses up to 1 0 0 %. In India, tungro disease was first identified in 1967
but acquired epidemic proportions in 1973 and 1981 in the north-eastern
region. The disease is transmitted by rice green leafhoppers Nephotettix
virescens and Nephotettix nigropictus. These leafhoppers acquire the virus
while feeding on phloem, of the infected rice plants, and transmit the
same to healthy plants during feeding.
Etiology
The virus is a nonpersistent leafhopper transmitted virus. It can readily
be transmitted within a minimum period of two hours after acquisition.

190 Rice Breeding and Genetics; Research Priorities and Challenges
Quick and efficient transmission of the virus, occurrence of the vector in
vast numbers, easy and rapid build-up of the vector under early
vegetative growth under field conditions, quick moving and longdistance
migrating prolific breeding vectors spread the disease rapidly.
Tungro is endemic to areas with overlapping or continuous crop
planting. Off-season, the virus survives on wild rices, rice stubble, and
some weeds as symptomless carries.
The leaves of the affected plants turn orange or brick-red, coupled
with chlorosis in newly emerged leaves. The infected plants are greatly
stunted and may have a reduced number of tillers, and may not bear
panicles. Even if panicles emerge, they are reduced in length and bear
discolored and chaffy spikelets.
The symptoms differ based on strain and particle type of the rice
tungro vims (RTV). Rivera and Ou (1967) first reported the existence of
strain sin RTV from the Philippines. Two strains designated as "s" and
"m" were identified. Anjaneyulu and John (1972) subsequently
identified four strains: RTV 1, RTV 2A, RTV 2B, RTV3. Misra et al (1976)
added RTV 4 to the list. It is known that RTV is a complex made up of
separate viruses, which can be separated on the basis of symptomology,
transmission characters, and the electron microscopic structure of the
virus particles itself.
Genetics
No thorough analysis of the mode of inheritance of tungro resistance has
been carried out to date. Preliminary reports indicate that two genes
may convey resistance in some varieties. A study of the Pankhari 203/
Taichung Native 1 at IRRI indicated that resistance in Pankhari 203 is
governed by two complementary dominant genes (IRRI, 1967).
According to Shastry et al, (1972), resistance in Latisail is under duplicate
gene control giving a segregation ratio of 9 : 7.
After compilation of the International Rice Tungro Virus Project,
Ling et al (1981) reported that the reaction of 1 0 rice varieties, viz. TNI,
IR26, Ambemohar 159, Habiganj, DW8 , Kataribhog, Latisail, Pankhari
203, IR34, Gam Pai 30-1-2-15, and Ptb 18 differed in different localities,
indicating strainal variation of RTV. They further concluded that
Habiganj, DW 8 , and Ptb 18 can be used to separate the RTV strains.
Some accessions of wild rice species, e.g. O. grandiglumis, O. latifolia, O.
malampzhuaensis, O. minuta, O. officinalis, O. perrieri, and O.
schweinforthiana, have been reported to be resistant to RTV (Anjaneyulu
et al, 1981). A number of improved plant type varieties have been
developed using one or the other donors (Tables 9.2 and 9.3),
i ¡Ml'

Ram C. Chaudhary 191
Breeding
The earliest and classic work on breeding for tungro resistance was
carried out in Indonesia in the early 1930s. Mentek disease, now known
to be identical to Tungro, has caused serious crop losses in the country
since 1859. Even though its exact nature was not exactly understood at
that time, several varieties were found resistant to it when grown in
mentek-infected areas. At Bogor in 1934, Van der Muelen crossed the
resistant variety Latisail from India with the susceptible variety Tjina
from China. Seeds from the F4 bulk of that cross were divided into four
parts and grown at four experiment stations located in the regions
where mentek was causing serious crop losses. Three stations were in
West Java and the fourth in East Java. Individual plant selections were
made at each station on the basis of resistance to mentek. Single plant
progeny lines were further evaluated for mentek resistance, yielding
ability, and agronomic traits at the stations and in farmers' fields. From
the trials, several advanced generation breeding lines with mentek
resistance and good agronomic traits were identified and released as
varieties. Those selected were Bengawan (at Ngale in East Java), Peta,
Mas, and Intan (at Singamerta), Fadjar (in West Java), Pelopor and Salak
(Bogor), and Tjahaja (at Tjitajam, near Bogor). Intan and Mas were
released for commercial production in 1940; Tjahaja, Fadjar, Pelopor,
Bengawan, and Peta, in 1941; and Salak, in 1942. Within a short time, the
varieties were distributed all over the country, where they gradually
replaced the old, susceptible varieties, particularly where mentek was
known to be serious. Varieties Peta, Intan, Bengawan, and Mas became
especially popular in Indonesia, the Philippines, Thailand, Bangladesh,
and India where all four proved resistant to tungro (Gangopadhyay and
Padmanabhan, 1987).
The breeding program for tungro resistance at IRRI was started
during 1966-1967. Several resistant varieties—Peta, Intan, Sigadis,
TKM6 , HR21, Malagkit Sungsong, Gam Pai, Ptb 18, Pankhari 203, and
BJl—^were donor parents. Improved plant type breeding lines with
tungro resistance were identified from the crosses of most of those
parents (Table 9.3) Seven IRRI-named varieties are moderately to highly
resistant to tungro. IR 20, IR 26 and IR 30 inherit their moderate
resistance from TKM6 . Gam Pai 15 is the donor parent of the highly
resistant IR 28, IR 29, and IR 34. The IRRI breeding program for
resistance was expanded and screening under field conditions started in
1971, Tungro resistance is one of the major breeding objectives in
Indonesia, Malaysia, the Philippines, Thailand, Bangladesh and India.
Variety C4063 developed in the Philippines, RD5 in Thailand, and BR41
M

192 Rice Breeding and Genetics: Research Priorities arid Challenges
Bangladesh are moderately resistant. In India Triveni is highly resistant,
and Vijaya, Ratna, and Pusa 2-21 are moderately resistant.
Several resistance donors, including Pankhari 203, Ptb 18, Gam Pai,
TKM6 , and HR 21, have been utilized in the resistance breeding program
of IRRI since 1966-67. Most IR varieties, except IRS, IR8 , IR22,
IR24, and IR43 have varying levels of tungro resistance.- Varieties IR28,
IR29, IR34, IR50, IR52, IR54, IR56, IR58, and IR60 are highly resistant.
Varieties IR32, IR36, IR38, IR42, and IR48 are moderately resistant. The
rice improvement program of Indonesia, Malaysia, Thailand, the Philippines,
Bangladesh, and India incorporate resistance to RTV in the improved
varieties.
Table 9.3
Some improved plant typ>e varieties, tungro-resistant varieties, and
selections developed at IRRI from various donor parents.
Donor parent Variety / Selection Cross
TKM6 IR20 Peta/TN1//TKM6
IR26
IR24/TKM6
IR30
IR24/TKM6/IR20VO. nivara
Gam Pai 15 IR28 PetaVTNl/Gam Pai 15/4/IR8/
Tadukan//TKM6 ^
/TN1////IK24VO. nivara
IR29
Peta^ /TNI/Gam Pai 15/4/IR8/
Tadukan//TKM6 ^
/TN1////IR24VO. nivara
IR34
Peta^ /TNI/Gam Pai 15/4/IR8/
Tadukan//TKM6 *
/TN1////IR24^ / 0. nivara
Ptb 18 IR32 IR20^ / 0 . nivara//CR94-13
1R36 IR8/Tadukan/TKM6VTN1 ///
IR24^ / 0 . nivara
/4/CR94-13
HR21 IR2034-289-1 IR24//Mudgo/IR8///Peta3 /TNI
//HR21/4/IR24VO. nivara
Pankhari 203 IR825-11-2 IR8 /Pankahri 203//Peta® / TNI
Sigadis IR127-80-1 CP231/SIO 17//Sigadis
Grassy Stunt
Agati et al. (1941) described from the Philippines for the first time the
symptoms of a disease that appeared to be grassy stunt. The disease first
appeared on the IRRI farm in 1963 and its transmission by the brown
planthopper (Nilaparvata Ingens) was demonstrated in 1964 (Rivera et al,
1966). Grassy stunt has now been reported from Thailand, Sri Lanka,
Indonesia, India, Bangladesh, Vietnam, China, Cambodia, Laos,
Myanmar (Khush, 1977a). When fully developed, symptoms on the
diseased plants are expressed as severe stunting, excessive tillering, and

Ram C. Chaudhary 193
an erect habit. The leaves turn short; narroW; erect, pale green or pale
yellow, and often have numerous small dark brown spots of various
shapes, which may form blotches. The leaves may turn green when
supplied with adequate nitrogenous fertilizers (Ling, 1972), Depending
on the age of the plant at which it is infected, the yield losses may range
from 0% to 100% (Ling, 1972).
Etiology
Since the reactions of varieties from various countries are identical, it
was postulated that there is no strainal variation in the virus (Khush,
1977a). Of late, virologists at IRRI Philippines have identified a new
strain of the virus based on serological relationship and similarity in
morphology, symptomatology, virus-vector interaction. This new strain
has been designated as rice grassy stimt 2 (RGS V2 ) versus the original
strain RGS VI. RGS V2 infected plants show stunting, leaf yellowing and
spreading growth habit. However, the symptoms may vary depending
on the variety and the age of the plant. Leaves of some varieties are
mottled or striped and have irregular rusty blotches. Plants infected at
the seedling stage show profuse tillering and narrow leaves, as do plants
infected with RGS Vl^ and die prematurely. However, plants infected at
later growth stages develop symptoms indistinguishable from those
caused by the tungro virus infection. Symptoms of this type are most
prevalent in the field (Hibino et at, 1983).
RGS V2 differs from RGS VI in pathogenicity to rice varieties. O.
nivara, which is resistant to RGS VI, is susceptible to TGS V2.
Consequently, all the varieties with O. nivara resistance are susceptible
to RGS V2 , which is prevalent in the Philippines, Thailand and
Indonesia (Hibino et aï., 1983).
Inheritance
A single dominant gene Gd confers resistance to grassy stunt disease.
INSECT PESTS
Rice grows in hot-humid environments where insect pests also flourish
and damage crops. More than 100 species of insects are considered rice
pests but only 20 species are of major economic importance. These
species infest all parts of the rice plant at one or the other growing stage.
But host-plant resistance is available only against a limited number of
insect pests. Genetics of resistance has been reported against these pests
and systematic breeding has been undertaken, resulting in the release of
varieties resistant to these pests.

194 Rice Breeding and Genetics: Research Priorities and Challenges
Brown Planthopper
The brown planthopper, Nilaparvata lugens (Stal) is the most serious of
the rice pests. The distribution of brown planthopper (BPH) in various
rice-growing countries has been listed by Flint and Magor (1982) and
Pathak and Khan (1994). Due to, host specificity a number of biotypes of
BPH developed in due course of time. Biotypes 1 and 2 are widely
distributed in southeast Asia. Bio type 3 is a laboratory biotype produced
in the Philippines and biotype 4 occurs in the Indian subcontinent.
D amage
ilil
BPH causes considerable damage by direct feeding. BPH sucks the sap
and plugs the xylem and phloem with its feeding sheath and pieces of
tissue pushed into these vessels during exploratory feeding. This direct
feeding may result in "hopper burn" which results in 100% crop loss. In
addition to the damage by feeding, it also transmits grassy stunt and
ragged stunt viral diseases, which may cause 50-90% losses due to
resultant high sterility and panicle deformation.
; {i
i
G enetics
More than 100 resistant cultivars have been genetically analyzed.
Athwal et al (1971) showed that the resistance in Mudgo, C 022, and
MTU15 was governed by the same dominant gene, which they
designated Bph-1. A single recessive gene, designated bph-2, conveyed
resistance in ASD7. Bph-1 and bph-2 are closely linked and no
recombination between them has been observed. Chen and Chang (1971)
also reported that a single dominant gene controls resistance in Mudgo.
Athwal and Pathak (1972) reported that MGL2 possesses Bph-1, and Ptb
18 possesses bph-2. Martinez and Khush (1974) investigated the
inheritance of resistance in two breeding lines of rice that originated
from the crosses of susceptible parents. One of the lines, IR747B2-6,
possessed Bph-1 , and the other, IRl154-243, possessed bph-2. It was
hypothesized that one of the parents, TKM6 has Bph- 1 and also an
inhibitory gene I-Bph-1 while some progenies segregate for these two
genes due to independent recombination and give resistant descendants.
In a genetic study of 28 varieties, Lakshminarayana and Khush
(1977) found 9 varieties with Bph-1, lis with bph-2, and one variety with
both genes. Two varieties were found to have new genes. A single
dominant gene, which conveys resistance in Rathu Heenati was
designated as Bph-3. This gene segregates independent of Bph-1. A
single recessive gene, which controls resistance in Babawee, was

Ram C. Chaudhary 195
designated bph~4. This gene segregates independent of bph-2. Genetic
analysis of 20 resistant varieties by Sidhu and Khush (1978) revealed
that 7 varieties had Bph-3/10 had bph-4/ and resistance in the remaining
three was governed by two genes. Sidhu and Khush (1979) also reported
that Bph-3 and bph-4 were closely linked. Genes bph-4 and GLH-3 were
also linked with a map distance of 34 units. The bph-4 gene appeared to
be linked with sd-1 (recessive gene for semidwarf stature). However^
bph-4 and Xa4 (gene for bacterial blight resistance) are inherited
independently. Ikeda and Kaneda (1981) also foimd that bph-2 as well as
Bph-1 segregate independent of both Bph-3 and bph-4, whereas Bph-3
and bph-4 as well as Bph-1 and bph-2 are closely linked. Ikeda and
Kaneda (1982) reported that Bph-1 segregated independent of the gene
for dwarf virus resistance in Kanto PL-3 as well as the gene governing
stripe disease resistance in Kanto PL-2.
On the basis of trisomie analysis, Ikeda and Kaneda (1981) identified
the loci of Bph-3 and bph-4 on chromosome 10. In another study, Ikeda
and Kaneda (1983) located Bph-1 on chromosome 4, No linkage was
detected between Bph-1 on the one hand and Ig and d-11 markers on
chromosome 4 on the other. However, bph- 2 was found linked with d-2,
with a 39.4% recombination value. Khush et al. (1985) carried out a
genetic analysis of ARC10550. This cultivar is resistant to BPH
populations in Bangladesh and India (biotype 4), but is susceptible to
biotypes 1 ,2 , and 3. It was found to have a single recessive gene, bph-5,
for resistance, which segregates independent of Bph-1,. bph-2, Bph-3,
and bph-4.
Seventeen additional rice cultivars, resistant to biotype 4 but
susceptible to biotypes 1, 2 and 3 were genetically analzyed by Kabir
and Khush (1988). Seven were found to have a single dominant gene for
resistance. The dominant gene(s) of these cultivars segregated
independent of bph-5. Thé dominant gane of Swarnalata was designated
as Bph-6 . In the remaining 10 cultivars, resistance is conferred by a
single recessive gene. The recessive genes of eight, cultivars were found
to be allelic to bph-5. However, the recessive genes of two cultivars are
nonallelic to bph-5. The recessive gene of T12 was designated bph-7.
Two Thai varieties. Col. 5 Thailand and Col. 11 Thailand, and Chin
Saba from Myanmar were reported to have single recessive genes,
which are allelic to each other but are non allelic to bph-2 and bph-4.
Similarly cultivars Kaharmana, Balamawee, and Pokkali were found to
have single dominant genes that are allelic to each other but different
from Bph-1 and Bph-3 (Ikeda, 1985). Since these cultivars are resistant to
bio types 1,2, and 3, compared to cultivars with bph-5, Bph- 6 and bph-7,
which are susceptible, Nemoto et al (1989) concluded that the recessive
gene of Col. 5 Thailand, Col. 7. Thailand, and Chin Saba must also be
1
!

196 Rice Breeding and Genetics: Research Priorities and Challenges
different from bph-5 and bph-7. They designated this gene as bph-8 .
Similarly, they designated the dominant gene of Kajharmana,
Balamawee and Pokkali as Bph-9. Several genes for resistance to the
BPH have been transferred from wild Oryza species to cultivated rice
through wide hybridization (Jena and Khush, 1990). Genetic analysis to
determine allelic relationships of these genes with known genes is
underway. An introgression line from the cross of cultivated rice and O.
australiensis has a dominant gene for BPH resistance, which has been
designated as Bph-10. Bph-10 confers resistance to three biotypes to
BPH. The genes for resistance in rice varieties can be inferred without
genetic analysis by determining their reaction to different biotypes
(Table 9.4). Most of the varieties released for resistance to BPH also have
multiple resistance to various diseases and pests (Tables 9.2 and 9.5).
Breeding
i' i
Bph“l confers resistance to Biotypes 1 and 3, and bph-2 against biotypes
1 and 2. Bph-3 and bph-4 confer resistance to all known biotypes. The
genes bph-5, Bph-6 , and bph-7 confer resistance to Biotype 4 only, and
bph- 8 and Bph-9 provide resistance to biotypes 1, 2, and 3 (Table 9.5),
Based on these considerations, systematic breeding for resistance against
BPH has been done in all the countries where this pest causes serious
damage, These resistance genes provide most of the protection to the
rice varieties.
Sources of resistance to BPH were indetified in 1967 (Pathak ei al,
1969). The pagoram on breeding and genetics was started in 1968. Two
genes for resistance, Bph-1 and bph-2 , were identified in 1970 (Athwal et
al., 1971). The first resistant variety with Bph-1, IR 26, was released in
1973 (Khush, 1977a). The variety was widely accepted in the Philippines,
Indonesia, and Vietnam but became susceptible in 1976-1977 because of
the development of biotype 2 of the BPH. By that time, varieties IR 36
and IR 38 with bph- 2 gene had been developed and released (Khush,
1977b), IR 36 soon replaced IR 26 and became the dominant rice variety.
Its resistance to BPH has held up for 14 years in most areas and it is still
widely grown (Table 9.4). Most of the varieties released for resistance to
BPH, also have multiple resistance to various diseases and pests
(Table 9.5).
Incorporation of newly identified genes Bph-3 and bph-4 continued
after their identification. In 1982, when a biotype capable to damaging
IR 36 appeared in small pockets of the Philippines and in Indonesia,
IR 56 and IR 60 with Bph-3 for resistance were released (IRRI, 1983). IR
6 6 with bph-4 for resistance was released in 1987 and IR 6 8 , IR 79, IR 72
and IR 74. All with Bph-3-were released in 1988. These varieties are now
widely grown in tropical and subtropical rice-growing countries. Thus,

J i
Ram C. Chaudhary 197
the continuous identification of new genes followed by their
incorporation maintained the genetic diversity of resistance genes and
helped keep ahead of the shifting enemy, BPH.
White-backed Planthopper
White-backed planthopper (WBPH) Sogatella furcifera (Horvdth) is one of
the major pests of rice in South and Southeast Asia, Pacific region, and
Australia.
D a m a g e
Like the brown planthopper, WBPH likewise causes considerable
damage by direct feeding. WBPH sucks the sap and plugs the xylem and
phloem with its feeding sheath and pieces of tissue pushed into these
vessels during exploratory feeding. Direct feeding may result in "hopper
burn" which results in 100% crop loss. Fortunately WBPH is not known
to transmit any viral diseases.
Table 9.4 Interrelationships between biotypes of BPH and gene for resistance in rice
Variety Gene Chromosome,
location
Reaction to biotype
1 2 3 4
Mudgo Bph-1 1 2 R S R S
ASD7 bph-2 4 R R S S
Rathu Heenati Bph-3 1 0 R R R R
Babawee bph-4 10 R R R R
ARC 10550 bph-5 - S S S R
Swarnalata Bph-6 - S S S R
T 1 2 bph-7 - s s S R
Chin Saba bph-8 - R R R -
Balamawee Bph-9 - R R R -
0 . australiensis Bph-lOm 1 2 R R R R
TN (1) None - S S S S
R : Resistant; S : Susceptible.
Table 9.5
Current knowledge of varietal resistance against green leafhopper,
zigzag leafhopper, brown planthopper and white-backed planthopper
Insect- Resis- Chro- Reference Varietal source of resistance
pest tance mosome
(1)
gene
(2 )
location
(3) (4) (5)
Green Glh-1
-
Athwal et al., 1971 Pankhari 203
leafliopper
Glh-2 Athwal et al., 1971 ASD7
Clh-3 10 AthwaUfaL 1971 IR8
glh-4 - Siwi and Khush 1977 P tb 8
(Contd.)

198 Rice Breeding and Genetics: Research Priorities and Challenges
Ml
:| Il
M!
Table 9.5 Contd.
Insect- Resis- Chro- Reference Varietal source of resistance
pest tance mosome
gene location
(1) (2 ) (3) (4) (5)
Glh~5 - Siwi and Khush, 1977 ASD 8
Qlh-6 5 Rezaul Karim and Pathak, TAPL # 796
1982
Glh-7 —■ Rezaul Karim and Pathaky Moddai Karuppan
1982
Gih-8 -
DV 85
Zigzag Zlh-l “ Angeles et al., 1986 Rathu Heenati
leafhopper
Zlh-2 - Angeles et al., 1986 Ptb21
Zlh-3 - Angeles et al., 1986 Ptb 33
Brown Bph~l 12 Sidhu and Khush, 1978 Mudgo
planthopper
Bph-2 4 Sidhu and Khush, 1978 ASD7
Bph-3 1 0 Lakshminarayana and Rathu Heenati
Khush, 1977
Bph-4 1 0 Lakshminarayana and Babawee
Khush, 1977
Bph-5 - Khush et al., 1985 ARC 10550
Bph-6 - Kabir and Khush, 1988 Swarnalata
Bph‘7 - Kabir and Khush, 1988 T12
Bph-8 . - Nemoto et aL, 1989 Chin Saba
Bph-9 Nemoto et al., 1989 Pokkali
Bph- 12 Khush et al., 1994 O, australiensis
10(t) ..
White-backed Wph-1 7 Sidhu et ai, 1979 N22
planthopper
Wph~2 . - Angeles etal., 1981 ARC 10239
Wpk-3 “ Hernandez and Khush, ADR 52
1981
Wph-4 — Hernandez and Khush, Podiwi AB
1981
Wph-5 Wuand Khush, 1984 N'Diang Marie
Wph-6(t) - Min et al., 1981
Gall Midge Gm-1 Chaudhary et ah, 1986 W1263,
Eswarakora
Gm-2 4 Chaudhary et al. 1986 Phalguna, Siam
29, Leuang 152
gm-3 - Sahii et al., 1985 RP 2068-18-3-5
Gm-4 (0 - Srivastava et al., 1994 Abhaya
Gm-5 (0 - Srivastava et al., 1994 ARC5984
Gm-6 (t) - Katiyar et al., 1995 Duokang # 1
Gm-7(t) - Reddy etaL, 1997 Bhumansan
Gm-8 (t)
NHTA8
Gm-9 (t) - NHTA8
Gm~10(t) - NHTA8
Gm-11 (t)
Banglei
G e n e t ic s
More than 300 cultivars resistant to WBFH have been identified and 80
of them have been analyzed genetically. Five genes for resistance^ one

Ram C. Chaudhary 199
recessive and four dominant^ have been identified. A single dominant
gene^ designated Wbph-1^ was found to convey resistance to the WBPH
in the variety N22 (Sidhu et ah, 1979). Resistance in ARC 10239 is
governed by a single dominant gene designated Wbph-2 (Angeles et al.,
1981). This gene segregates independently of Wbph-1. Nair et al. (1982)
investigated 21 additional varieties: 19 had Wbph^l and two had Wbph-
2 and an additional recessive gene. The resistance of 2 of the 14 varieties
analyzed by Hernandez and Khush (1981) was governed by Wbph-2.
Eleven varieties had a single dominant gene each that segregated
independent of Wbph-1 and Wbph-2. The dominant gene of one such
variety, ADR52, was designated Wbph-3. Only one variety, Podiwi A 8 ,
had a recessive gene, which was designated wbph-4. Saini et al. (1982)
analyzed 13 additional varieties. Resistance was governed by Wbph-1 in
four varieties, Wbph-2 in six, Wbph-1 and Wbph-2 in two, and a single
dominant gene in Hornamawee segregated independent of Wbph-1 and
Wbph-2. Wu and Khush (1985) investigated the inheritance of resistance
in 15 varieties. They found that resistance in nine varieties was
controlled by Wbph-1, and two genes conferred resistance in four
varieties. The remaining two varieties had single dominant genes for
resistance, which segregated independent of Wbph-1, Wbph-2, and
Wbph-3. The dominant gene of N'Diang Marie was designated Wbph-5.
Jayaraj and Murty (1983) studied the inheritance of resistance in nine
varieties. They found that it was controlled by a single dominant gene in
three varieties and by a recessive gene in six other varieties.
Inheritance of resistance in 10 cultivars was investigated by Singh et
al. (1993). Eight cultivars, i.e., ARC 5838, ARC 6579, ARC 6624, ARC,
ARC 10464, ARC 11321, ARC 11320, Balamawee, and IR2425-90-4-3,
were found to have single recessive genes for resistance. The recessive
genes of IR2415-90-4-3, ARC 5838, and ARC 11324 were found to have
single recessive genes for resistance. The recessive genes of IR2415-90-4-
3, ARC 5838, and ARC 11324 were found to be allelic to each other.
Resistance in Ptbl9 and IET6288 was found to be under dominant gene
control.
Green Leafhopper
Several species of green leafhopper (GLH) are rice pests but three are of
economic significance. Nephotetix cincticeps (Uhler) distributed in China,
Japan, and Korea is a vector of rice dwarf and yellow dwarf. Nephotetix
virescens (Distant) distributed in South and Southeast Asia is a vector of
yellow dwarf, tungro, penyakit merah, and yellow-orange leaf.
Nephotetix nigropictus is also distributed in South and Southeast Asia

200 Rice Breeding and Genetics: Research Priorities and Challenges
and is a known vector of rice dwarf, yellow dwarf, transitory yellowing,
tungro, yellow-orange leaf, and rice gall dwarf.
Damage
Other than the feeding damage done by sucking, which results in
reduced growth of the crop at a young stage, GLH is also .a vector of a
number of viruses, which once spread can cause serious yield losses.
G enetics
Athwal et ah (1971) first investigated the inheritance of resistance to
GLH in varieties Pankhari 203, ASD7, and IR 8 . They found that
resistance in each variety was controlled by one dominant gene. The
dominant gene in Pankhari 203 was designated Glh-1; that in ASD7,
Glh-2; and that in IR 8 , Glh-3. The three genes segregated independent of
each other. Siwi and Khush (1977) identified two more genes; one
recessive designated glh-4 and the other dominant designated Glh-5.
Two dominant genes, Glh- 6 and Glh-7, were identified by Rezaul Karim
and Pathak (1982).
Aveshi and Khush (1984) studied the inheritance of resistance in 18
varieties. Two had Glh-1, three had Glh-2 , two had Glh-3, one glh-4, and
three had two genes. The allelic relationships of the resistance genes of
seven varieties are still not known. Ruangsook and Khush (1987)
analyzed 15 rice cultivars genetically. The resistance was governed by
two dominant genes in Katia Badger 13-20, Laki 659, Lasane, Asmaita,
and Choron Bawala, but by single dominant genes in the remaining 10
cultivars. Later it was found that one of the two dominant genes of
Choron Bawla is allelic to Glh-2. The single dominant gene in Chiknaql
and one of the two dominant genes of Laki 659 are allelic to Glh-3. The
second of the two dominant genes of Katia Badger 13-20, Laki 659, and
Lasane are allelic to Glh-5. The two dominant genes of Asmaita and the
single dominant gene of Hashikalmi, Ghaiya, ARC 10313, and Garia are
nonallelic to and independent of Glh-1, Glh-2, Glh-3, glh-4, and Glh-5.
Tomar and Tomar (1987) studied the inheritance of resistance in 11
cultivars. Resistance in eight cultivars was found to be governed by
single dominant genes, while single recessive genes conferred resistance
in the three other cultivars. Inheritance studies of resistance in 12
cultivars by Ghani and Khush (1988) revealed single dominant genes in
six cultivars, two independent dominant genes in four cultivars, and
single recessive genes in two other cultivars. The single recessive gene in
ARC 7012 is allelic to glh-4 but that in DV85 is nonallelic to and independent
of glh-4. The recessive gene was designated glh-8 .

Ram C. Chaudhary 201
Breeding
A large number of donors have been used to breed resistant varieties
against GLH. Earlier varieties released by IRRI Philippines/ e.g. IR5/ IRS,
IR20; IR24/ IR26, IR30 had the Glh-3 gene. Other genes, e.g. glh-4 and
Glh-5, have also been used. As a result of continuous breeding effort to
introduce more than one gene into the newly developing varieties, a
large number of varieties and lines have become available with a good
degree of resistance. Some of the varieties bred by IRRI using various
resistance genes are given in Tables 9.2 and 9.5.
Zigzag Leafhopper
The zigzag leafhopper, Recilia dorsalis {Motschulsky) is so known not
because of its damage pattern but rather the zigzag coloring pattern on
the wings of adults. The zigzag leafhopper (ZLH) is distributed
throughout South Asian and Southeast Asian countries.
Damage
The zigzag leafhopper, Recilia dorsalis (Motschulsky) is a leafhopper not
so much known for serious damage due to sucking of plant sap but as a
vector of serious viral diseases such as gall dwarf, tungro, and yelloworange
leaf virus.
Genetics
The genetics of resistance to the zigzag leafhopper (ZLH) was
investigated by Angeles et at (1986) in such cultivars as Rathu Heenati,
Ptb21, and Ptb33. Single dominant genes that segregate independent of
each other and conveyed resistance to ZLH damage were designated
Zlh-1 (Rathu Heenati), Zlh-2 (Ptb21), and Zlh-3 (Ptb33) are. Tests for
independence of various genes for resistance to leaf and planthoppers
revealed that Zlh-1, Zlh-2, and Zlh-3 are independent of Wbph-3. Zlh-2
and Zlh-3 also segregated independent of bph-2 and Bph-3.
Gall Midge
The rice gall midge, Orseolia oryzae {Wood-Mctöon) is a serious pest of rice
in certain areas of South and Southeast Asia. It has been reported from
Bangladesh, China, India, Indonesia, Laos, Myanmar, Nepal, Pakistan,
Sri Lanka, Thailand, and Vietnam. In Africa, another species of gall
midge, Orseolia oryzivora (Harris and Gagne), damages the crop, but is not
a serious pest. It is reported to occur in Cameroon, Ghana, the Ivory
Coast, Liberia, Mali, Niger, Nigeria, Senegal, and Sudan. Both species

H.-:
202 Rice Breeding and Genetics: Research Priorities and Challenges
require high humidity and thus lowland rices are damaged more than
upland.
Damage
The Asian rice gall midge (GM) Orseolia oryzae {Wood-Mason) is a serious
pest in many countries of South and Southeast Asia, causing 10-100%
yield losses. The characteristic symptom of attack is a long, hollow,
tubular gall commonly called a "silver shoot", a response to a secretion
of the salivary glands containing "cecidogen", which causes
proliferation of cells at the site of feeding. The larvae feed on the
growing point of the tiller and the growing tiller is converted into a
tubular gall. Tillers with silver shoots do not produce panicles and dry
off. In addition, GM induces the prolonged tillering, delays flowering,
and reduces the number of ears bearing tillers and 1 , 0 0 0 grain weight as
well as yield. The existence of differential biotypes based on differential
varietal reaction has been known for some time.
African rice gall rpidge is yet another pest but with a similar mode of
damage that occurs in several African countries. Only limited
information on the inheritance of varietal resistance against it is
available.
Genetics
In India, six biotypes have been characterized to date (J. Bentur, pers,
comm.). Gall Midge biotype (GMB) 1 is confined to central Andhra
Pradesh, Madhya Pradesh and parts of Orissa. GMB 2 is confined to
coastal Orissa, while GMB 3 is distributed in Northern Telengana of
Andhra Pradesh and southern Bihar. Biotype GMB 4 is widely distributed
in the northern coastal areas of Andhra Pradesh and Vidarbha
of Maharashtra. GMB 5 has limited distribution in Kuttanad area of
Kerala, GMB 6 is distributed in Manipur and is believed to be present in
other neighboring countries.
Resistance to gall midge has been postulated to be due to two genes
in W1263 and four genes in Ptbl8 (Shastry et aL, 1972). Sastry and
Prakasa Rao (1973) inferred the presence of three recessive genes for
resistance in W1263 and W12708. Satyanarayaiah and Reddi (1972),
however, convincingly showed that resistance in W1263 is governed by
a single dominant gene. Two to three dominant complementary genes
for resistance (Sastry et ah, 1984) govern resistance in CR57-MR-1523.
Chaudhary et al. (1986) studied the inheritance of resistance in five
cultivars, all of which were found to have a single dominant gene for
resistance. Allele tests revealed that Usha, Samridhi, W1263 and BD6-1
have the same gene for resistance, which was designated Gm-1. Surekha,
Phalguna and IET6285 have the same gene for resistance, which is

Ram C. Chaudhary 203
nonallelic to and independent of Gm-1; this gene was designated Gm-2,
Kalode et al. (1976) found different reactions of W1263 and JBS 446 at
two locations, indicating biotypic variation in gall midge. The gene gm-
3 (t) was reported from variety RP 2068-18-3-5. Variety Abhaya was
released with yet another gene, designated Gm-4(t). The gene Gm-5 (t)
was reported from variety ARC 5984, and the gene Gm- 6 (t) from
Duokang No. 1. Variety Bhumansan was reported to have two genes,
Gm-7 (t) and Gm- 8 (t). Similarly, NHTA 8 was reported to have two
genes, namely Gm-9 (t) and Gm-lO(l). Recently, the gene Gm- 1 1 was
reported from variety Banglei.
Breeding
The existence of biotype and its geographic distribution, complexity of
reaction, and damage has made breeding of resistant varieties difficult.
Still a number of varieties have been developed that are resistant to a
number of biotypes.
Planting a resistant variety is the most effective means of preventing
gall midge damage. Differences in varietal susceptibility to this pest
were reported as early as 1922 in Vietnam, and in 1927 in India. Several
rice improvement programs of Bangladesh, India, Indonesia, Sri Lanka,
and Thailand are currently screening varieties and have a regular
breeding program for resistance. In India, several rice varieties, such as
Eswarakora, HR42, HR63, Ptb 18, Ptb 21, Siam 29, and the Thai variety
Leung 152, were found highly resistant. Resistance to gall midge is
reported to be primarily due to antibiosis. Larval development is
retarded on resistant varieties but is normal on susceptible varieties.
Genes Gm-2, Gm-4(t), and GM6 -(t) have been tagged to molecular
markers and thus it should be possible to start maker-aided selection.
This will also solve the problems of screening varieties against all the
biotypes of gall midge, as their collection at one location is forbidden Q.
Bentur, pers. comm.).
Stem Borer
The stem borers of rice belong to the order Lepidoptera, principally the
families of Pyralidae and Noctuidae. Thirty-five pyralids belonging to
12 genera, 10 noctuid species belonging to 3 genera, 5 diopsid species
belonging to the genus Diopsis have been recorded as rice stem borers
(Chaudhary et ah, 1984; Pathak and Khan, 1994). Five of these—^yellow
borer {Scirpophaga incertulus Walker), striped borer {Chilo suppressalis
Walker), white borer {Scirpophaga innotata Walker), dark-headed borer
{Chilo polychrysus Meyrick), pink borer {Sesamia inferens Walker)— are of
economic significance in Asia. The yellow borer is primarily distributed

204 Rice Breeding and Genetics: Research Priorities and Challenges
in the tropics, but also occurs in temperate areas where temperature
remains above 100“C and rainfall above 1,000 mm. In African countries,
stalked-eyed fly (Diopsis macrophthalma)' and African white-heads
{Malirpha separatella) are the major stem borer species causing extensive
damage. In the American continent. South American white borer {Rupela
albinella) is the major stem borer species.
DAMAGE
iNl
In Asia, the yellow borer and the striped borer are the major pests and
widely distributed from India to Japan. They are reported to be
responsible for a steady annual damage of 5-10% of the rice crop with
local catastrophic outbreaks of 60% damage (Jepson, 1954). All the
species, except the pink borer, which lays eggs between the leaf sheath
and the stem instead of on the leaf blade, have the same mode of life
history. Eggs hatch after one week and within one or two days firstinstar
larvae migrate to between the leaf sheath and the stem where they
feed on the sheath. During the second instar, they bore inside the stem
where they feed inside the lumen. Pathak (1968) described the damage
caused by borers as follows: the initial boring and feeding by the larvae
in the leaf sheath causes broad longitudinal whitish discolored areas at
the feeding sites, but rarely results in wilting and drying of the leaf
blades. About a week after hatching, the larvae cease feeding on the leaf
sheath and bore into the stem and feed on the inner tissue of stem walls.
Such feeding frequently results in severing of the apical parts of the
plant from the point of damage. When this kind of damage occurs
during the vegetative phase of the plant, the central whorl of leaves does
not unfold, turns brownish and dries, while the lower leaves remain
healthy and green. This condition is known as "dead-heart" and the
affected tillers die without bearing panicles. Larvae feeding above the
primordia some times cause "dead-hearts" but if no further damage
occurs, the severed portions get pushed out by new growth.
After panicle initiation, severing of the growing plant parts from the
base results in drying of the panicles. These panicles may not emerge at
all. Those that do emerge produce no grains and remain straight and
appear whitish; they are called "white-heads" When the larvae at the
base damage the newly emerged panicles, partially filled grains are
formed. Plants can compensate for a low percentage of "dead-hearts"
but for every percent of "white-heads" 1-3% of yield loss may be
expected (Pathak et at, 1971). Although stem-borer damage becomes
evident only as "dead-hearts" or "white-heads", yet significant losses
result from the feeding of larvae within the stem without severing the
growing point (Catling and Islam, 1981) and reduction in plant vigor
and yield.

Ram C. Chaudhary 205
Genetics
The available donors only have a moderate degree of resistance (Table
9.6) on the standard scoring system (INGER, 1995) Under field
conditions^ a number of scapes occur resulting into misclassification of
susceptible plants as resistant. An additional complication in inheritance
studies occurs ,due to various types of resistance in different varieties^
starting from morphological to anatomical to biochemical and antibiosis.
Therefore^ conclusive studies on the inheritance of resistance to stem
borer resistance are hard to come by.
Apparently, the only information on genetics of resistance available
earlier than 1964 (Pathak, 1964) was that of Koshiary et al. (1957) who
crossed Giza 14, a resistant variety, with Sydney A. Based on the study
of p2 plants and F3 progenies, they suggested the resistance to be under
polygenic control. The polygenic nature of inheritance was also reported
in variety Chianan 2 in crosses with Rexoro (Pathak, 1970; IRRI, 1973).
Athwai and Pathak (1972) investigated the inheritance of resistance to
the striped borer by studying the body weight of larvae fed on the F2
plants and the dead-heart" counts in Rexoro/TKM 6 cross. Resistance,
i.e., low body weight, was dominant in F^; in F2 about 75% larvae had
low body weight. Thus it was concluded that resistance was a simply
inherited trait. But the "dead-heart" counts of the F2 progenies in the
field did not show a definite pattern of segregation although the
reaction was dominant; thus it was concluded that several genes for
resistance might be involved.
Based on the logic of increasing the level of resistance in breeding
lines, Khush (1977b) concluded that the gene action for striped borer
resistance was additive.
Breeding
During the last 40 years, large numbers of germplasm have been
screened against various species of stem borers. Details for various
species, screening done in various countries, etc. are reviewed by
Chaudhary et al. (1984), and Khan et al. (1991). Extensive screenings
have been done for resistance against striped borer. A generalized
scoring system, which is widely followed (INGER, 1995) for scoring
resistance, has been given in Table 9.6. Most varieties now being
developed and released have a good degree of resistance and even
multiple resistance (see Table 9,2).
International Rice Research Institute (IRRI)
Work on breeding for resistance of the striped borer at IRRI was initiated
in 1966. Several donors were crossed with improved plant type
germplasm. Many improved breeding lines with resistance similar to or

206 Rice Breeding and Genetics: Research Priorities and Challenges
Table 9.6 Scoring system of damage due to stem borer (based on SES, INGER, 1995).
Score Dead-hearts % White-heads % Description
0 0% 0 % Highly resistant
1 1- 1 0% 1-5% Resistant
3 1 1 - 2 0 % 6- 1 0 % Moderately resistant
5 21-30% 11-15% Moderately susceptible
7 31-60% 16-25% Susceptible
9 61% and above 26% and above Highly susceptible
U
better than that of their donor parents including multiple disease and
insect resistance were developed. Because the donors were related, it
was assumed that they had different genes for resistance. Based on this
assumption, a program was initiated in 1972 to accumulate genes for
different sources to develop improved germplasm by diailel selective
mating (DSM) as proposed by Jensen (1970). The method involves: (1)
crossing a number of moderately resistant parents in all possible
combinations; (2 ) intercrossing the Fj populations so obtained in all
possible combinations; (3) screening the double crossed Fj progeny for
resistance; (4) intercrossing the plants found, to have better resistance
than either of the parents. The crossing, screening, selection and
recrossing are continued until minor genes from different sources are
accumulated and the intensity of the trait is built up. The DSM is a type
of recurrent selection involving a broad gene pool, breaking linkages,
freeing of genetic variability and fostering of genetic recombination.
Striped borer: Breeding for resistance to striped borer at IRRI started
in 1965 when IR 8 and Feta *3/TN l were crossed with TKM. By 1968 a
number of progenies, e.g. IR532E239, IR532E257, IR532E576 (released as
IR20 later) with superior resistance and yield had been identified. A
number of crosses such as IR7474, IR1330, IR1514A, and IR1561 were
made to incorporate striped borer resistance into good agronomic
background (IRRI, 1971). In the following years, a number of progenies
from multiple crosses of moderately resistant lines were screened (IRRI,
1972, 1973, 1974), and several breeding lines from progenies of IR1702,
IR2031, IR2058, IR2061, IR2070, IR2071, IR2151, and IR2153 crosses were
identified as having a good degree of resistance. Some of the promising
lines were released as IR28, IR32, and IR34 (IRRI, 1975, 1976). One
breeding line, IR4442-207-2-3, was found to have the best level of
' resistance apparently combined from TKM6 and CR94-13,
A DSM system was started in 1972 to accumulate genes for resistance
from seven different breeding lines, namely CR94-13, IR5-156
(Mehran 69), IR1365-83-2-5-3, IRl 416-131-5-2-3, IR1514A-E666, IR1561-
228-3-3, and IRl721-11-13-25. Another group of crosses were made
using the same scheme, and three cycles of selective mating produced
superior donors (IRRI, 1976).

Ram C. Chaudhary 207
Yellow borer: The breeding program for yellow borer resistance
started after 1972 when screen-house techniques for screening were
developed and some donors were identified. Three improved plant type
lines-IR1721-ll/ IR1917-3 and IRl820-52-2—were found resistant. IR
1820-52-2 had a higher level of resistance than all the previously known
donors though none of its parents were known for their resistance (IRRI^
1976). IR34 was released as a moderately resistant variety in 1975^ and
moderately resistant lines IR2071-625-1-252 and IR2070-414-3-9 were
released by the Philippine government as IR36 and IR40 respectively in
1976. Such efforts were also reported later.
A new breeding approach to upgrade the level of stem borer resistance
was adopted in 1980^ using the male-sterility-facilitated-composite
(MSFC) and male-sterility-facilitated-recurrent-selection (MSRS) scheme
(Chaudharyj, 1981). Genetic male sterile IR36ms was used as the female
parent to cross with 26 known donors of yellow borer resistance and a
composite population was created by mixing the seed of each single
cross in equal amoiints. The composite was grown during the 1982 dry
season in such a way so as to synchronize the maximum tillering phase
with the peak population of yellow stem borer. It was assumed that
under heavy pressure only those plants with a good degree of resistance
would come to heading and contribute pollen for seed setting on the
male sterile plants. Only the seeds set on the male sterile plants were
harvested to start the next cycle (Chaudhary and Khush, 1990).
India
Beginning in 1964/ TKM6 / CBl, and CB2 were used as resistant donors
and a large number of crosses were made with semidwarf high-yielding
varieties as well as with local improved lines (Roy et aL, 1971). Stem
borer resistance trials were started in 1968 with entries emanating from
TKM6 / IR532/ and Eswarakora and continued until 1975 with newer
(AICRIP 1969/ 1975) entries from such donors as Ptbl8 / ptb21/ and
Eswarakora, But TKM appeared in the percentage of 85.9% entries.
Jagannath was the first resistant variety released in the post-IR8 era. It
was followed by a series of resistant entries: Cauvery/ Ratna, Vijaya,
Pusa 2-21/ Hamsa/ Rajendra/ Phalguna/ Supriya/ Kumar/ Parijat/ Saket 4
in various states of India (Seetharaman and Sobha Rani/1979).
Other Countries
Varietal development for stem borer resistance depended in several
countries primarily on introductions. In Bangladesh/ breeding lines such
as IR20/ Chandina/ (IR532-1-176)/ Purbachi (Chen-chu Ai I)/ IR28/ lET
2845/ lET 5540, and IR580 were introduced, international testing of the
known resistant lines and varieties commenced in 1976 in the form of
the International Rice Stem Borer Nursery (IRSBN) with 6 8 entries. The

208 Rice Breeding and Genetics: Research Priorities and Challenges
trial was continued for over a decade and served a useful purpose of
testing and making the resistant donors globally available (INGER1995),
i
i
M a n a g e m e n t o f s t e m b o r e r u s in g m o d e r a t e l y r e s is t a n t v a r ie t ie s
Since none of the commercially available varieties have a very high level
of resistance, management of the stem borer using moderately resistant
varieties appears to be a bright possibility. The importance of moderate
level of resistance was demonstrated in a study of the population growth
of the striped borer at IRRI (IRRI, 1972). IR20 and Taitung 16 were used
as the resistant varieties and Rexoro as the susceptible one. Taitung 16
has an antibiosis type of resistance. Fewer larvae survived on Taitung 16
than on ÏR20 and Rexoro (Table 9.7), Survival on Taitung 16 declined
with each subsequent generation. The larvae raised on IR20 and Taitung
16 were smaller than those on Rexoro (Table 97), Consequently, the
moths reared on these two varieties were smaller and laid fewer eggs.
Also, larvae reared on these two varieties produced fewer female moths.
Although the varieties had the same number of larvae at the start, the
fourth generation moths produced 82 times more eggs on Rexoro
compared on Taitung 16. Combined effects of these factors produced
much lower borer population (IRRI, 1972).
Table 9.7
Population growth of the striped stem borer (SSB) on selected rice varieties,
(susceptible) Rexoro, (resistant) 1R20 and Taitung 16 (IRRI, 1972).
Rice
variety
Generation
of
SSB
Survival
(%)
Avg. wt(mg)
Larva
Pupa
Days for
50%
pupation
Sex
ratio
Egg
masses/
female
Eggs/
mass
Rexoro 1 72.0 87.4 27 m 5.1
2 - 68,3 63.9 30 1.09 3.5 19.1
3 ■68.9 63.1 71.6 27 1.17 4.9 23.9
4 69,8 85.9 71.4 28 1.16 4.2 2 2 .8
IR20 1 69.7 - 43.5 26 1.09 4.8
2 - 48,6 46.0 30 1.06 2.4 14.1
3 67.0 45,2 47.0 29 1.06 2 .8 16.4
4 64.8 54.2 47.6 29 1.06 2.9 15.4
Taitung 1 49.2 30.8 32.5 30 1.08 •4.7 19.3
16 2 52.0 34.3 31.3 31 1.09 2 .8 15,6
3 41.6 28.5 40 1.09 2.7 13.3
4 17.5 25.7 42 1.09 2.5 11.3
FUTURE DIRECTION
The next millenium will witness some of the toughest challenges in rice
supply for the growing populace. The present production has to be

Ram C. Ghaudhary 209
doubled in the next 10 years using less land, water, labor and chemicals
(fertilizers, pesticides, fungicides, bactericides, etc.). The socioeconomj-c
situation and environmental consciousness will impose these
production constraints. Above all, stability in the production has to be
added on a sustainable basis. The sum total of the given scenario dictates
that host-plant resistance will have to receive top priority.
On the positive side, newer tools provided by biotechnology and
related developments will make the identification and transfer of more
potent genes easier. Some of the developments are briefly described
below.
(a) Gene pools for potent genes: Traditionally, the rice gene pool
consists of wild species, weed races> land races, cultivars, breeding
stocks, released varieties, and induced mutants. All these types of
germplasm can be considered as primary and secondary types (Harlan
artd De Wet, 1971). The primary gene pool consists of about 150,000
accessions of land races, varieties, weed races, and 4 wild species, viz.
0 . nivara, O.rufipogan, O. barthii, O. longistaminata (Table 9.8). The secondary
gene pool consists of O. punctata, O. officinaliis, O. eichingeri, O.
minuta, O. latifolia, and O. hrachyantha, Th.e tertiary gene pool consists of
more distantly related species such as 0 . meyeriana, O. ridleyi, and
species belonging to related genera such as Leersia hexandra, Porteresia
coarctata which are difficult to cross using conventional mean’s. In the
not too distant future, these could be targeted using biotechnological
tools already available.
(b) Conventional breeding methods: While no major shift is expected
in the traditional breeding methods such as pure line selection, mass
selection, hybridization and selection using the pedigree method, bulk
method, single-seed descent, back cross and mutation, emphasis will be
placed on a few novel ones. Breeding methods such as male-sterilityfacilitated-recurrent
selection (Chaudhary et ah, 1981; Chaudhary and
Khush, 1990), wide hybridization (Brar and Khush, 1991), markerraided
selection (Tanksley, 1983), and hybrid rice breeding will hold center
stage (Yuan, 1993; Virmani, 1994). Newer tools such as somaclonal
variation and genetic engineering (Khush, 1998) will find more
prominent use.
(c) Wide hybridization: Wild species are more resistant or sometimes
immune (Devadath, 1983) to various insect pests and diseases but
the transfer of resistance genes to cultivated rice posed problems
through conventional means (Khush and Brar, 1991). But methods
became available whereby a number of very useful genes have been
transferred from wild species to cultivated rice (Table 9.8). Jena and
Khush, 1990, Multani et ah, 1994). Such transfers will speed up with the
identification more resistant genes and potent tools to aid the transfer.

210 Rice Breeding and Genetics: Research Priorities and Challenges
Table 9.8
Genes for disease ai\d insect resistance transferred from wild species
into cultivated rice
Alien donor Genome Resistance to Reference
Oryza australiensis EE Bacterial blight and BPFi Ishii et ah, 1994
Oryza bracHyantha FF Yellow stem borer Khush etal., 1994
Oryza latifolia CCDD Bacterial blight, BPH
Amante ei ah, 1992
and tungro
Oryza longistaminata Bacterial blight Devadath, 1983
Oryza minuta BBCC Brown planthopper
Amante et ah, 1992
and sheath blight
Oryza nivara AA Grassy stunt virus resistance Khush, 1977
Oryza officianlis c c BPH, WBPH and tungro Jena and Khush, 1990
Oryza ridleyi Unknown Tungro and yellow stem borer Khush et ah, 1994
Í'
1
Ü
; : 1
: 1
■¡': ' !'
hi- !^:
IT
1T
liir
iI :
J ^!
(d) Marker-aided selection: Marker-aided selection helps in th
cases in which the resistance gene under reference is tightly liked with
plant morphological traits so that the resistance and the morphological
trait segregate together in the breeding population. Plants selected on
the basis of a morphological trait are automatically resistant. Such
situations are seldom encoimtered, however. But the molecular markers
(isozymes and restriction fragment lengths polymorphisms, IRFLPs) are
very useful (Mackill and Bonman, 1992) for tagging genes of economic
importance. They differ from morphological markers in several respects
(Tanksley, 1983):
(i) The genotypes of molecular loci can be determined at the whole
plant, tissue, and cellular levels. The phenotypes of most
morphological markers can only be distinguished at the whole
plant level.
(ii) A relatively large number of naturally occurring alleles is found
at molecular loci. Distinguishable alleles at morphological marker
loci occur less frequently.
(hi) Usually deleterious effects are associated with alternate alleles of
molecular markers. Undesirable phenotypic effects, on the other
hand, often accompany morphological markers.
(iv)
possible genotypes to be distinguished in any segregating
generation. Alleles at morphological marker loci usually interact
in a dominant-recessive manner, prohibiting their use in many
crosses.
(V) With morphological marker loci, strong epistatic effects limit the
number of segregating markers that can be unequivocally scored
in the same segregating generation. With molecular markers, very
few epistatic or pleitropic effects are observed.
If
markers can be tagged (Table 9,9), selection efficiency can be increased

Ram C. Chaudhary 211
and the time and money spent on selections can be minimized. The
presence or absence of the associated molecular marker would indicate^
at a very early stage^ the presence or absence of the desired gene.
Co dominance of the molecular marker allows all possible genotypes to
foe identified in any breeding scheme even if the economic gene cannot
be scored directly. A tight linkage (less than 5 cM) is necessary for
tagging a gene with a single molecular marmer. For a gene-tagging
approach to be successful, molecular markers placed at. intervals of less
than 5 cM throughout the genome are required. Isozyme markers are
not numerous enough in any crop to inark the whole genome. However,
RFLP markers are numerous and saturated maps can be prepared
(BChush et al, 1994). In a number of cases, resistance to insects is
quantitative and is governed by polygenes (Khush et al., 1994). In such
c^ses, it is important to establish linkage between the molecular markers
and the polygenes (quantative trait loci or QTL) to help in the selection.
(e) Genetic engineering: Tools of genetic engineering are most potent
to transfer resistance genes from incompatible crosses and unrelated
genera or species or even synthetic genes. Transgeiiic plants may be
produced using tools such as agrobacterium-mediated transformation,
electroporation, biolistic method, gene gun or projectile bombardment,
and microinjection. The Bt gene has already been transferred in rice
(Khush, 1998) and the transfer of a number of novel genes is still
underway. The Bt gene is known to be effective in controlling the striped
stem borer and the leaf folder.
Table 9.9
List of resistance genes in rice tagged with molecular markers
Gene
Resistance
to
Donor variety
Chromosome
location
Reference
Bph-10 (t) BPH 0 . australiensis 12 Ishii et al, 1994
Gm-2 Gall midge Siam 29 4 Mohan etal, 1994
Hbv Hoja blanca Fanny 12 see Khush et al, 1994
Pi-2 (t) Blast LAC 23 11 Xu etal, 1991
Pi-2 (t) Blast 5173 6 Yu et al, 1991
Pi-4 (t) Blast Tetep 12 Yu et al, 1991
Pi-?(t) Blast IRAT13 4 CIAT, 1991
Pi-sm Blast Moroberekan 4 Wang, 1994
Pi-zh Blast Zhaiyeqing 8 see Khush et al, 1994
Wph-1 WBPH N 22 7 McCouch, 1990
Xa-1 BB Kogyku 4 Yoshimura et al, 1992
Xa-2 BB Tetep 4 Yoshimura etal, 1992
Xa-3 BB Chogoku45 11 Yoshimura et al, 1992
Xa-4 BB IR20 11 Yoshimura et al, 1992
xa-5 BB IR1545-339 5 McCouch et al, 1991
Xa-W BB CAS 209 11 see Khush et al, 1994
Ka-21 BB O. hngistamimta Ronald et al, 1992

212 Rice Breeding and Genetics: Research Priorities and Challenges
!íi í'
References
Adair, C.R. 1991. Tech. Bull USDA 171.
Aga ti, J.A., Sisón, P.L. and Abalos, R. 1941. Philipp.}. Agrie 12:197-210.
AICRIP (All India Coordíanted Rice Improvement P r o j e c t , 1 9 6 9 , Report, Vols. landll,
ICAR, New Delhi, India.
AICRIP {All India Coordinated Rice Improvement Project) 1971, Ann. Rep. Vols. I, III, ICAR,
New Delhi, India.
Akai S., Ogura, H. and Sato, T. 1960. Annals Phytopath. Soc. Jpn. 25:125-130.
AICRIP (All India Coordinated Rice Improvement Project 1975, Annual Report. Vols I and III,
ICAR, New Delhi, India.
Anjaneyulu, A. and John, V.T. 1972. Phytopath. 62; 1116-1119.
Angeles, E.R., Khush, G.G. S. and Heinrichs. E. A. 1986. In: Rice Genetics, IRRI, Los Baños,
Philippines, pp 537-549.
Anjaneyulu, A., Shukla V.D., Rao, G.M, and Singh, S.K. 1981. Ini. Rice Res. Newslett. 6(1): 12-
13,
Angeles, E.R., Khush, G.S. and Heinrichs, E.A, 1981. Crop Sei. 6:551-554.
Anon, 1995. R.ice Genet. Newslett. 12:9-154.
Athwal, D.S. and Pathak, M.D. 1972. In. Rice Breeding IRRl, Los Baños, Laguna, Philippines
pp. 375-386,
Athwal, D.S., Pathak, M.D., Badangco, E.H, and Pura, C.D. 1971, Crop Sei. 11:747-750.
Atkins, J.G. and Johnston, T.H. 1965. Phytopatk. 55:993-995.
Aveshi, G.M. and Khush, G.S. 1984. Crop Prot. 3:41-52.
Balal, M.S., Selim, A.K., Hassavien, S.H. and Maximoos, M. A. 1977. Egyptian }. Gen. Cytol 6;
332-341.
Bonman, J.M., Khush, G.S. and Nelson, R.J. 1992, Ann. Rev. Phytopath. 30:507-528.
Bonman, J.M., Estrada, B.A., Kim, C.K., Ra, B.S. and Lee, E.J. 1991. Plant Disease 75:468-466.
Brar, D.S. and Khush, G.S. 1991. In: Handbook of Plant Cell Culture, 4, McMillan Publ. Co., New
York, USA pp. 221-263.
Buddenhagen, I.W. 1983. In: Durable Resistance in Crops. Plenum Press, New York, USA. pp.
401-428.
Buddenhagen, I.W. and Reddy, A.P.K, 1972, In: Rice Breeding, IRRI, Los Baños, Philippines,
pp. 289-295.
Catling, H.D. and Islam, Z. 1981. Proc. Deepwater Rice Workshop, Bangkok, Thailand, IRRI,
Philippines.
Chang, T.T. 1962. IRC Newslett. 11 (2): 1-7.
Chang, T.T., Wang, M.K., Lin, K.M. and Geng, C.P. 1965. In: Rice Blast Disease, The Johns
Hopkins Press, Baltimore, Maryland, USA. pp. 371-377.
Chaudhary, B.P., Srivastava, P.S., Shrivastava, M.N. and Khush, G.S. 1986. In: Rice Genetics
IRRI, Los Baños, Laguna, Philippines, pp. 523-528.
Chaudhary, R.C. and Nayak, P.R. 1987. In; Advances in Rice Pathology, TNAU-ICAR, New
Delhi, India pp. 1-35.
Chaudhary, R.C. and Khush, G.S. 1990. Insect Sei. and Applic. 11:659-669.
Chaudhary, R.C., Heinrichs, E.A. and Khush, G.S. 1981. /«f. Rice Res. Newslett. 6:7-8.
Chaudhary, R.C., Khush, G.S. and Heinrichs, E.A. 1984. Insect Set. Applic. 5(6): 447-463.
Chen, L.C. and Chang, W.L. 1971. /. Taiwan Agrie. Res. 20:57-60.
CIAT (Centro Internactional de Agricultura Tropical) 1991. Cali, Colombia, pp. 51^6.
Correa-Victoria, F.J., Zeigler, R.S., Tohme, J. and Levy, M. 1992. Proc. Symp. Durability
Disease Resistance. lAC, Wageningen, Netherlands.

Ram C. Chaudhary 213
Crill, P„ Ham, Y,S, and Beachell, H.M. 1981, Korean J. Breeding 13(2): 106-114.
Devadath, S. 1983. Current Sei. 52:27-28.
Dye, D.W,, Bradbung, J.F., Goto, M., Hayward, A.C., Lelbiatt, R.A. and Shroth, M. N. 1980.
Rev. Plant Pathol. 59:153-169.
Ezuka, A, 1979, Proc. Rice Blast Workshop, IRRI, Los Baños, Philippines, pp. 27-48,
Bzuka, A. and Harino, K. 1974. Bull Tokai-Kinki Natl. Agrk. Exp. Stn. 27:19-31.
Flint, P. and Magor, J.I. 1982. ODA Mise, Rep. No. 58, Centre for Overseas Pest Research,
London.
Fujimaki, H., Kiyosawa, ,S. and Yokoo, M, 1975. Ann. Phytopath. Soc. Japan 41:176-184.
Gangopadhyaya, S, and Padmanabhan, S.Y. 1987. Breeding for Disease Resistance in Rice.
Oxford & IBH Publ, New Delhi, India 340 pp.
Ghani, M.U. and Khush, G.S. 1988. /. Genet. 67:151-159.
Goita, M. 1985. Dies. Abs. Int. B. 46(5): 1808B.
Goto, 1.1970. Ann. Phytopath. Soc., Japan 36:304-312.
Goto, 1.1976. Ann. Phytopath. Soc., Japan 42:239-244,
Goto, 1.1978. Ann. Phytopath. Soc., Japan 44:447-455,
Goto, 1.1979. Ann. Phytopaih. Soc., Japan 45:252-255.
Goto, 1.1988. Ann. Phytopath. Soc., Japan 54:460-465.
Goto, I,, Jaw, Y.L. and Baluch, A.A. 1981. Ann. Phytopath. Soc., Japan 47:252-255.
Harlan, J.R. and De Wet, J, M.J. 1971. Taxon 20:509-517.
Hashioka, Y. 1950. Taiwan Agrie. Agio. Res. Inst. Tech. Bull. 8; 237.
Hashloka, Y. 1951. Ann. Phytopatk. Soc. Japan, 15:98-99
Hernandez, J.E. and Khush, G.S. 1981. Oryza 18(1): 45-50.
Hibino, H., Tinongco, E.R. and Cabunagan, R.C. 1983. Int. Rice Res. Newslett. 8(4); 12-13.
Hsieh, S.C. and Chang, S.C. 1977. Plant Prot. Bull. Taiwan, 19(1): 275-286.
Ikeda, R. 1985. Bull. Natl. Agrie. Res. Cent. 3:1-54.
Ikeda, R, and Kaneda, C. 1981, Jap. J. Breed. 31; 279-285.
Ikeda, R. and Kaneda, C. 1982. Jap. J. Breed. 32:177-185.
Ikeda, R. and Kaneda, C. 1983. Jap. J. Breed, 33:40-44.
Imbe, T. and Matsumoto, S. 1985. Inheritance of resistance of rice varieties to blast fungus
strains virulent of the variety "Reiho". Jp«, J. Breed. 35:332-359 (In Japanese with English
Summary).
INGER 1995. Proc, Advisory Committee Meeting, IRRI, Los Baños, Philippines, 345 pp.
Inukai,T., Nelson, R.J., Mackill, D.J., Bonmañ, J.H., Takamar, I., and Kinoshita,T. 1996. Theor.
Appl. Genet. 93:560-567.
Inukai, T., Nelson, R.J., Zeigler, R.S., Sarkarung, S., Mackill, D.J., Bonman, J.M., Takamar, I,
and Kinoshota, T. 1994. 84:1278-1283.
IRRI (International Rice Research institute) 1967. Ann. Rep. 1966. IRRI, Los Baños, Laguna,
Philippines, 302 pp.
IRRI 1969. Ann. Rep. 1970. IRRI, Los Baños, Laguna, Philippines, 402 pp.
IRRI 1971. Ann. Rep. 1970. IRRI, Los Baños, Laguna, Philippines, 266 pp.
IRRI 1972. Ann. Rep. 1971 IRRI, Los Baños, Laguna, Philippines, 265 pp.
IRRI 1973. Ann. Rep. 1972, IRRI, Los Baños, Laguna, Philippines, 246 pp.
IRRI 1974. Ann. Rep. 1973, IRRI, Los Baños, Laguna, Philippines, 266 pp.
ÍRRI1975. Ann. Rep. 1974, IRRI, Los Bafios, Laguna, Philippines, 384. pp.
IRRI 1976. Ann. Rep. 1975. IRRI, Los Baños, Laguna, Philippines, 479 pp.
IRRI 1983. Ann. Rep. 1982. IRRI, Los Baños, Laguna, Philippines.
IRRI 1984. Ann. Rep. 1983. IRRI, Los Baños, Laguna, Philippines.

214 Rice Breeding and Genetics: Research Priorities and Challenges
IIHH'
Ishii, T., Brar, D.S, Multam, D.S. and Khushy G,S. 1994. Molecular tagging of genes for brown
planthopper resistance and earliness introgressed fvomOryza australiensis into cultivated
rice, O. Sativa. Genome 37: 217-221.
Ito, R. 1965. In: The Rice Blast Disease. The John Hopkins Press, Baltimore, Maryland, USA pp,
361-^70.
Iwata, K. and Omura, T, 1971. }ap. f. Breed. 21:16-17.
Jayaraj, D, and Murty, V.V.S. 1983. Froc. 15th Int. Cong. Genet. 1983. AbstractNo. 1309, p. 724.
Jena, K.K. and Khush, G.S. 1990. Theor. Appl. Genet. 80: 737-745.
Jensen, N.F. 1970; Crop Sci. 10:629-635. .
Jepson, W.F. 1954. Commonwealth Inst, of Entomology, London.
Johnson, R. and Bonman, N.M. 1993. In: New Frontiers in Rice Research, DRR, Hyderabad,
India pp. 206-212.
Kabir, M.A. and Khush, G.S. 1988. Plant Breed. 100:54-58.
Kalode, M.B., Shastry, M.V.S. and Pophaly D.J. 1976. J. Biol. Scii 19:62-65.
Kaneda, C. 1980. Gamma Field Symp. 27: 71-89.
Kannaiyan, S. and Prasad, N.N. 1978. ÍMíer, Rice Res. Newslett. 2(1): 6.
Katiyar, S.K., Tan. Y,, Zhang, Y., Huang, B., Xu, Y,, Zhao, L., Huang, N,, Khush, G.S. and
Bennett. 1995. Molecular tagging of gall midge resistance genes in rice. In: Fragile Lives in
Fragile Ecosystems. IRRI Los Baños, Philippines, pp. 935-948.
Kaur, S., Kaur, P. and Padmanabhan, S.Y. 1984. Indian Phytopath. 37:100.
Khan, Z.R., Litsinger, J.A., Barrion, A.T., Villanueva, F.F.D., Fernandez, N.J. and Taylo, L.D.
1991. World Bibliography of Rice Stem Borers. IRRI, Los Baños, Philippines.
Khush, G.S. 1977a. Ann. N.Y. Acad. Sci. 28:296-308.
Khush, G,S, 1977b. Adv. Agron. 29: 265-341.
Khush, G.S. 1989. In; Progress in Irrigated Rice Research IRRI, Los Baños, Laguna, Philippines,
pp. 79-82.
Khush, G.S. 1998. Proc. 3rd Asian Crop Science Conf. Taichung, Taiwan roc. Abstract, p. 28.
Khush, G.S. and Brar, D.S. 1991. Adi?. Agron. 45:223-274.
Khush, G.S., Rizaul Karim, A.N.M. and Angeles, E.R. 1985. /. Genet. 64:121-125.
Khush, G.S., Brar, D.S., Zapata, F.J., Nelson, R., McCouch,S. and Bottrell, G. 1994. In; Toward
Enhanced and Sustainable Agricultural Productivity in 2000 s, Vol. II. Taichung District Agrie.
Imp. Stn., Taiwan roc, pp. 387-414.
Kiyosawa, S. 1967. Genetic studies on host-pathogen relationship in rice blast disease, Proc.
Symp. Rice diseases and their control by governing resistant varieties and other
measures, Tokyo, pp. 137-153.
Kiyosawa, S. 1968. Jap. J. Breed, 18; 193-205,
Kiyosawa, S. 1969. Jap, J. Breed. 19; 121-128.
Kiyosawa, S. 1970. Bull, Natl. Inst. Agrie. Sci. D 21:73-105.
Kiyosawa, S. 1972a. Bull. Natl. l«sf. Agrie. Sci. D 23:69-96.
Kiyosawa, S. 1972b. Jap. J. Breed. 12:140-146.
Kiyosawa, S. 1974a. Agrie. Hortic, 49: 445-446.
Kiyosawa, S. 1974b. Jap. J. Breed. 24:117-124.
Kiyosawa, S. 1974c. Mise. Publ. Natl. Inst. Agri. Sci. 1; 1-58.
kiyosawa, S. 1976. Oryza 13:1-32.
Kiyosawa, S. 1978. Jap. J. Breed. 28: 287-296.
Kiyosawa, S. 1981. Oryza, 18; 196-203.
Kiyosawa, S. 1983. Agrie. Hortic. 10:1432-1439.
Kiyosawa, S, and Murty, V.V.S. 1969. Jap, J, Breed. 22; 1269-276
Kiyosawa, S. and Yokoo, M. 1970. Jap. J. Breed, 20:181-186.
Kiyosawa, S. and Yabuki, S. 1976. Agrie. Horfic. 51:571-572.

Ram C. Chaudhary 215
Kiyosawa, S. and Cho, C.1.1980. ]ap. ]. Breed. 30:73-82.
Kiyosawa, S., Orimoto, Y., Kun, He. Yung, and Ling, Z.Z. 1983. Oryza 20:216-222.
Koshiary, M.A., Pan, C.I., Hak, G.E., Zaid, I.S.A., Azizi, A., Hindi, C. and Masoud, M. 1957.
IRC Newslett. 6: 23-25.
Lakshminarayana, A, and Khush, G.S, 1977. Crop Sci. 17; 96-100.
Lee, P.N. and Rush, M.C. 1983. Plant Disease 67:829-832.
Li, M.P., Ni, P.C., Chen, Y.G.G., and Shen J.H. 1983. Acta Agronomica Sínica 9:173-179.
Librojo, V., Kauffman, H.E. and Khush, G,S. 1976. SABRAO /. 8; 105-110.
Ling, K.C. 1972. Rice Virus Diseflses, IRRI Los Baños, Philippines, 134 pp.
Ling, K.C., John, V.T., Rao, P.S., Anjaneyulu, A., Miah, M.S.A., Ghosh, A., Koesnang, S. and
Flores, Z.M. 1981. iizt. Rice Res. Newsíeíí. 6(1): 12
Lozano, J.C. 1977. Plant Disease Rep. 61: 644-648.
MacKenzie, D.R. 1979. In: Proc. Rice Blast Workshop. IRRI, Los Baños, Philippines pp. 199-
216.
Mackill, D.J. and Bonman, J.M. 1992, Phytopath. 82: 746-749.
Martinez, C.R. and Khush, G.S. 1974. Crop Sci. 14:264-267.
Me. Couch, S.R. 1990. Construction and application of a molecular linkage map of rice based
on restriction fragement length polymorphism (RFLP). Ph. D. Thesis, Cornell University,
Ithaca, NY, USA.
Me. Couch, S.R., Abenes, M.L., Angeles, E.R., Khush, G.S. and Tanksley, S.D. 1991. Molecular
tagging of a recessive gene Xa-5, for resistance to bacterial blight of rice. Rice Genet.
Newslett. 8:143-145.
Mevii, T.W. and Vera Cruz, C.M. 1977. Int. Rice Res. Newslett. 2(3): 8,
Mew, T.W. and Vera Cruz, C.M. 1979. Phytcfpath. 69:152-155.
Mew, T.W. and Khush, G.S. 1981. In: Proc. Int. Conf. Plant Pathogenic Bacteria Cali, Colombia,
pp. 504-510.
Mew, T.W., Vera Cruz, C.M., Reyes, R. and Zaragoza, B.Z. 1971. IRRI Research Paper Series
No. 39,8 pp,
Min, S., Wu, M., Lu, Z. and Tsai, G. 1981, The inheritance of economic traits in early Hsein
rice and its role in high yield breeding Scientia Agriculturea Sínica 2:31-37 (in Chinese),
Misra, A.P. and Chatterjee, A.K. 1963. Indian Phytopath. 16:275-281.
Misra, R.K., Mathur, S.C. and Chaudhary, R.C. 1976. In: Proc. Symp. Increasing Rice Yield in
Kharif. CRRI, Cuttack, India, pp. 95-113.
Mizukami, T. 1966. Japan Agrie. Res. Qtly. 1:6-11.
Mizuta, H. 1956, Associated Plant Protection. Kyushu, Japan, pp, 100-102.
Mohan, M., Nair, S., Bentur, J.S., Rao, U.P. and Bennett, J. 1994. RFLP and RAPD mapping of
rice GM2 gene that confuse resistance to Biotype 10 gall midge (Orseolia Oryzae) Theor.
Appl. Genet. 87: 782-788.
Mohanty, C.R. and Gangopadhyay, S. 1982. Ann. Phytopath. Soc. japan 48:5.
Mohanty, C.R. and Gangopadhyay, S. 1985, Indian Phytopath. 36:54-59.
Multani, D.S., Jena, K.K., Brar, D.S., detos Reyes, B.G., Angeles, E.R. and Khush, G. S. 1994.
Theor. Appl. Genet. 88:102-109.
Murty, V.V.S. and Khush, G,S. 1972. In: Rice Breeding, IRRI, Los Baños, Philippines, pp. 301-
305
Nagai, I, and Hara, S. 1930. Jap. J. Bot 5:41.
Nagai, K., Fujikami, H. and Yoku,M. 1973./flp«fi Agrk. Res. Qtly. 7:63-70.
Nagvi, N.I. and Chattoo, B.B. 1996. Rice Genet. Ill; 570-576.
Nawaz, M. and Kausar, A.J. 1962. Biologia Lahore 8:35-48.
Nelson, R.R. 1978. Ann. Rev. Phytopath. 16; 359-378.

216 Rice Breeding and Genetics: Research Priorities and Challenges
Nemoto, H., Ikeda, R. and Kaneda, C. 1989. }ap. J. Breeding 39:23-28.
Ogawa, T., Morinaka, T., Fuji, K, and Kumura, T. 197S,Ann. Phi/lppath. Soc. Japan 44; 137-141.
Olufowote, J.O., Khush, G.S. and Kauffman, H.E. 1977.■ Phytopath, 67: 772-775.
Ou, S.H. 1979. In: Proc, Rice Blast Workshop. IRRI, Los Baños, Philippines, pp. 81-137.
Ou, S.H. 1985. Rice Diseases (2nd ed.). Commonwealth Mycological Inst., Kew, Surrey, UK.
380 pp.
Ou, S.H. and Nuque, P.L. 1963. IRC Klewslett. 12(4): 30-35.
Padmanabhan, S.Y. 1965. Cwrrertí Sei. 34:307-308.
Padmanabhan, S.Y. 1973. Ann. Rev. Phytopath. 11:11-26.
Padmanabhan, S.Y. 1974, Final Tech. Rep. US-PL-480 Proj. CRRI, Cuttack, India.
Padmanabhan, S.Y,, Ganguly, D. and Chandwani, G.H. 1966. Indian Phytopath, 31:72-75.
Panda, N. and Khush, G.S. 1995. Host Plant Resistance to Insects. CABI-IRRI, Wallingford, UK,
431 pp.
Pathak, M.D. 1964. Int. Rice Comm. Newslett. 13; 15-21.
Pathak, M.D. 1968. Ann, Rev. Ent. 13:257-294.
Pathak, M.D. 1970. In: Concept of Pest Management N.C. State Univ, Raleigh, pp. 138-157.
Pathak, M.D. and Khan, Z.R. 1994. Insect Pests of Rice. IRRI, Los Baños, Philippines, ICIPE,
Kenya, 89 pp.
Pathak, M.D., Chang, C.H. and Fortuno, N.E, 1969. Nature 223:502-505.
Pathak, M.D., Andres, F., Galacgac, N. and Ramos, R. 1971. ÍRRÍ Tech. Bidl. 11: 69 pp.
Petpisit, V., Khush, G.S. and Kauffman, H.E, 1977. Crop Sei, 17:551-554,
Prakasa Rao, P.S. 1984. Cecidology Internationale 6 :9-17.
Premlatha Dath, A. 1985. Sheath Blight of Rice and its Management, Assoc. Publ. Co., New
Delhi, 129 pp.
Premlatha Dath, A. 1990. Sheath Blight Disease of Rice and its Management. Assoc. Publ. Co.
New Delhi, India, \29 pp.
Rani, N.S. and Satyanarayana, K, 1982. Int. Rice Res. Neioslett. 7(5): 7
Rao, P.S. and Kauffman, H.E. 1977. Phytopath. 55: 281-284.
Rath, G.C. and Padmanabhan, S.Y. 1972. Proc. Indian Acad. Set. 76:106.
Rath, G.C. and Padmanabhan, S.Y. 1973. //Riso 22(2): 123-129.
Reddy, A. P.K., Miah, S.A. and Mew, T.W. 1986. In: Progress in Rainfed Lowland Rice, IRRI, Los
Baños, Philippines, 253-262 pp.
Reitsma, J. and Schure, P.S.J. 1950. Contrih. Gen. Agrie. Res. Stn. Bogor 117:1-17.
Reys, R.C., Vera Cru2;, C.M., Aballa, T.C., Baraoidan, M.R. and Mew, T.M. 1982. In: Revised
Rice Production Manual, IRRI, Los Baños, Philippines.
Rezul, Karim, A.N.M. and Pathak, M.D, 1982. Genetic control of iron deficiency chlorosis
tolerance in rice, Rt. N 6:115-116.
Rivera C.T. and Ou, S.H. 1967. Plant Disease Rep. 51; 877-^881.
Rivera, C.T., Ou, S.H. and lida, T.T. 1966. Plant Disease Rep. 50:453-456.
Robinson, R.A. 1973. Rev. Plant Pathol. 52:483-501.
Ronald, P.C., Albano, B., Tabien, R:, Abenes, L., Wu, K., Me. Couch, S.R. and Tanksley, S.D.
1992. Genetic and physical analysis of the rice bacterial leaf blight resistance locus Xfl-22,
Mol. Gen. Genet. 236:113-120.
Roy, J.K., Israel, P. and Panwar, M.S. 1971. Oryza 8:129-134.
Ruangsook, B. and Khush, G.S. 1987. Crop Protec. 6; 244-249.
Sahu, V.N., Mishra, R., Chaudhary, B.P., Srivastava, P.S. and Srivastava, M.N. 1985.
Inheritance of resistance to gall midge in rice REN 7:118-121.
Saini, R.S., Khush, G.S. and Heinrichs, E.A. 1982. Crop Protect. 1 :289-297.
Sastry, M.V.S. and Prakasa Rao, P.S. 1973. Curr. Sei. 42: 652-653.

Ram C, Chaudhary 217
Sastry, M.V.S., Prasada Rao, U., Kalode, M.B. and Sain M. 1984. Indian J. Genet. 44:325-328.
Satyanarayanaiah, K. and Reddi, M.V. 1972,>4iid/ini Agrie. J. 19; 1-8.
Seetharaman, R. and Shobha Rani, N. 1979. Dir. Extn., Ministry of Agrie. Govt, of India 79 pp.
Shastry, S.V.S., Freeman, W.H., Seshu, D.V., Israel, P. and Roy, J.K. 1972. In; Rice Breeding,
IRRI, Los Baños, Philippines, pp. 353-365.
Sidhu, G.S. and Khush, G.S. 1978. Theor. Appl. Genet. 33:199-203.
Sidhu, G.S. and Khush, G.S. 1979. Euphytica 28:233-237.
Sidhu, G.S., Khush, G.S. and Madrano F.G. 1979. Euphytica 28:227-232.
Singh, L. 1993. In: Host Plant ^sistance to Insects. G.S.- Dhaliwal and V. K. Dilawari (eds.).
Kalyani Publ., New Delhi, pp. 31-78.
Singh, R.A. and Pavgi, M.S. 1969. Plant Disease Rep. 53:444-445,
Singh, R.J., Khush, G.S. and Mew, T.W. 1983. Crop Sd. 23:558-560.
Siwi, B.H. and Khush, G.S. 1977, Crop Sci. 29:17-20.
Skaguchi, S., Suwa, T. and Murata, N. 1968. Bh//. N«í/. Jnsi. Agrie. Sci. Japan 16:1-18.
Swaminathan, M.B. 1979. In: Proc. V Mel. Wheat Genetics Symp. New Delhi, India, 1978.
Srivastava, M.N., Kumar, A., Sri vasta va, S.K. and Sahu, R.K. 1994. A new gene for gall midge
resistance in rice. Rice Genet. Newslett. 10:79-80.
Tagami, Y, and Mizukami, T. 1962. Spec. Rep. Plant Disease and Insect Pests Forecasting Service
10,112 pp. Tokyo, Japan.
Takahashi, Y. 1965. In: The Rice Blast Disease. John Hopkins Press, Baltimore, Maryland, USA
303 pp.
Tanksley, S.D, 1983. Plant Mol. Biol. Rep, 1; 3-^.
Tomar, J.B. and Tomar, S.D. 1987. Plant Breed. 98:47-52.
Toriyama, K. 1975. In: Horizontal I^sistance to the Blast Disease of Rice. CIAT, Cali, Colombia,
pp. 65-100.
Uchimiya, H., Handa, T. and Brar, D.S. 1989, J, Biotech. 12:1-20.
Uchimiya, H., Toki, S. and Brar, D.S. 1991. Gamma Field Symp. No. 30, NIAR Ibaraki, Japan,
pp, 151-163.
Virmani, S.S. (ed.) 1994. Hybrid Rice Technology: New Developments and Future Prospects IRRI,
Los Baños, Philippines, 296 pp.
Wang, G.L., Mackill, D.J., Bonman, J.M. and Tanksley, S.D. 1994. Genetics 136; 1421-1434.
Wu, C.F. and Khush, G.S. 1985. Crop Sci. 11; 505-509.
Wu, S.Z., Hsu, S.M., Chen, F.K., Choi, L.C. and Liu, J.M. 1981. Int. Rice Res. Newsletts, 6(5): 6.
Yamada, T. 1984. Jap. J. Breed. 34; 181-190.
Yamazaki, Y. and Kiyosawa, S. 1966, Studies on inheritance of rice varieties to blast I.
Inheritance of Japanese varieties is to several strains of the fungus. Bull. Natl. Ins. Agr. Sci.
D. 14; 36-39 (In Japanese).
Yamazaki, Y. and Kosaka, T. (eds.) 1980. Blast Disease and Breeding for Resistant Varieties of
Rice. Tokyo, Hukyu-Sha 340 pp.
Yokoo, M., and Kiyosawa, S, 1970. Jap. J. Breed. 20:129-132.
Yokoo, M., and Kikuchi, F. 1977. Jap, J. Breed. 27:123-130.
Yoshimura, A., Mew, T.M., Khush, G.S. and Omura. T. 1983. Phytopath. 73:1409-1412.
Yoshimura, S., Yoshimura, A., Saito, A,, Kishimoto, N., Kawase, M., Yano, M., Nakagahra,
M., Ogawa, T. and Iwata, N. 1992. RFLP analysis of introgressed chromosomal segments
in three near isogenic lines of rice bacterial blight resistance genes Xa-1, Xd-3 and Xa~4.
Jap. f. Genet. 67:29-37.
Yuan, L.P. 1993. In: New Frontiers in Rice Research. DRR, Hyderabad, India, pp. 86-90.
Yu, Z.H., Mackill, D.J., Bonmann, J.M., Tanksley, S.D. 1991. Tagging genes for blast resistance
in rice via linkage to RFLP markers. Theor. Appl. Genet. 81:471-476.

10
Breeding for Adverse Soil
Problems in Rice
B.N. Singh*
INTRODUCTION
Since the release of IR 8 in 1966, world rice area increased by 2 2 million
ha, from 126 Mha to 148 Mha in 1992. The paddy production during the
same period doubled from 261 million tons to 528 Mt. This was mainly
due to adoption of the improved high-yielding varieties with tolerance
to biophysical stresses, and better input-responsive production
technology. With the growing world population, paddy production has
to be increased to 810 Mt by the year 2025 (Rosegrant et al., 1995). The
increase in production should be met through an increase in
productivity per unit of land and an increase in rice area. Good rice
lands in periurban areas are fast declining due to rapid urbanization and
industrialization. The increase in area to meet the demand of rice will
have to come from marginal lands and problem soil areas. Globally, 138
Mha of such land are available, of which 23 Mha are potential areas for
rice cultivation (Boje-Klein, 1986). These adverse soils are widely
distributed in arid, semiarid, subhumid, humid, and temperate regions
of the world. Broadly adverse soils are classified as saline, sodic, acid
sulfate, calcareous, and acid uplands. The saline soils can either be
inland or coastal. These soils are characterized by multiple nutrient
stresses. In addition to this there is a deficiency or toxicity of nutrients.
Lowland Rice Breeder, WARDA Bouake, Côte d 'Ivoire

2.20 Rice Breeding and Genetics: Research Priorities and Challenges
individual which adversely affects soil quality. The nutrient deficiencies
are zinc, phosphorus, iron, sulfur and silica. The nutrient toxicides are
aluminum, iron, boron, hydrogen sulfide, and managanese. These
problem soils can be exploited for rice cultivation through varietal
improvement and soil amendments. No variety can tolerate acute
nutrient deficiency or toxicity, but some can tolerate these partially
better than others. Varieties tolerant to problem soils will reduce the
cost of reclamation through amendments. Ikehashi and Pormamperuma
(1976), Ponnamperuma (1984), Chaubey and Senadhira (1994), De Datta
et al (1994), and Flowers and Yeo (1995) have done some reviews on this
topic.
SOIL STRESSES ’
Inland Salinity
Globally around 77 Mha of land are affected by salinity, of which 72%
have a light to moderate degree of salinity (Munns and Richards, 1998).
In these areas, the water table is high and underground water may be
saline or normal. These soils have a pH ranging from 7,0 to 8.3,
exchangeable sodium less than 15%, and electrical conductivity in
saturation extract (ECe) of more than 4 mmho cm'^ at 25®C in the top 50
cm. Four mmho cm”^ is the point beyond which rice yields decline
appreciably as the salt content increases (Maas and Hoffman, 1977), The
salt composition is mainly of chlorides and sulfates of Na, Ca, and Mg.
These soils are deficient in nitrogen, phosphorus, and zinc. At present
about 5 Mha of saline lands are cultivated to rice. Problems of salinity
are increasing due to poor drainage in canal irrigation systems. Mapping
of groundwater quality is essential for its suitability to irrigation (Gupta,
1994). In Pakistan, of a total 15 Mha of irrigated land, 11 Mha are salt
affected. In Iraq, 50% of the 3.6 Mha suffer from salinity, in India-26
Mha, Indonesia-15 Mha, China-7.6 Mha, Malaysia-4.8 Mha, Bangladesh-
4 Mha, and in Egypt-about 1 Mha. In Australia 20%, China 15%, and
Israel 13% of irrigated land is affected by salinity (Abrol, 1986; Gleick,
1993; Huang and Rozelle, 1993; Ghassemi et al, 1995; Kijne et al, 1998).
Coastal Salinity
Around 27 Mha of land in the humid zone of coastal areas are inundated
by sea water. Most of these lands are under mangrove vegetation. Areas
closer to the mouth of estuaries have more saline water inundation than
areas farther away from the sea. There are about 2.1 Mha of coastal

B.N. Singh 221
saline soils in India, which are mainly found in Gujarat, Orissa, and
West Bengal. In West Africa, 1.5 Mha of cultivable mangrove swamp are
affected by salinity, of which only 0.2 Mha are under cultivation (Jones,
1986). Coastal soil salinity is also increasing in the Nile delta in Egypt,
and the northern Senegal Kiver delta. In many coastal saline areas, the
soils are acidic and are potential acid sulfate soils. In the dry season, the
salinity effect increases due to sea-water intrusion. Waterlogging is
common in coastal saline areas in the wet season.
Sodicity or Alkalinity
Around 31 Mha land in semiarid and subhumid areas of South and
Southeast Asia, Africa, and Australia have a pH ranging from 8.3 to
11.0, exchangeable sodium more than 15%, and ECe less than 4 mmho
cm“^in the top 50 cm. Their salt composition is mainly carbonates and
bicarbonates of sodium and calcium. These soils are deficient in
nitrogen, phosphorus, zinc and iron, and have boron toxicity. These
soils do not have a high water table and water percolation is low. In
Pakistan, 9.4 Mha and in India, 2.5 Mha soils in the Indo-Gangetic plains
are affected by sodicity (Pormamperuma and Bandyopadhya, 1980).
Calcareous Saline Sodic
About 6 Mha are classified as calcareous saline sodic soils. They have
high calcium carbonates (3-38%) and pH ranging from 7.5 to 11.0.
Severe zinc, iron and phosphorus deficiency occurs in such soils. These
types of soils are common in Pakistan and India. In northern Bihar and
eastern Uttar Pradesh in India, over 200,000 ha of calcareous saline sodic
soils with up to 49% free CaCQs are found.
Acid Sulfate
Around 13 Mha soils in Indonesia, Vietnam, Cambodia, Thailand,
Bangladesh, West Africa, and Venezuela have a pH ranging from 3.0 to
4.5, and are classified as acid sulfate. Around 507o of such lands are
under cultivation and the rest are potential area for rice cultivation. In
the associated mangrove, where sea-water immdation is reduced during
the dry season, acid sulfate soils develop over a period of time due to
sulfide or sulfuric acid deposition. These soils show iron and aluminum
toxicity and phosphorus deficiency.

222 Rice Breeding and Genetics; Research Priorities and Challenges
Feat Soils
Peat soils cover more than 200 Mha worldwide, of which 32 Mha are in
the tropics, specifically 22 Mha in Asia, 7 Mha in Latin America, and 3
Mha in Africa. In Indonesia, about 16.5 Mha peat soils occur in coastal
lowland, of which only 0.5 Mha is cultivated. The pH of peat soils ranges
from 3.5 to 7.5 and organic matter more than 65% by weight in a
minimum depth of 50 cm (Driessen, 1978). Substantial areas under peat
soils occur in Malaysia, Vietnam, Thailand, India (Kerala) and are being
reclaimed for cultivation, Rice grown in peat soils (Histosols) suffers
from N, P, K, Cu, and 2n, and Mo deficiencies.
Add Uplands
In tropical humid forests (Ultisols and Oxisols) of Latin America, West
Africa, and high rainfall areas in Meghalaya, India, upland rice is grown
on aerobic soils with low pH (4.0-6.5). Around 1,400 Mha of such lands
are accounted for in Latin America alone, and 300 Mha are potential
areas for upland rice (Sarkarung, 1986). Availability of P in these soils is
reduced by reaction of soluble P with iron and aluminum oxides. These
soils in general suffer from P, K, Al, Ca, Mg, and Zn, and Si deficiencies.
Phosphorus (P) Deficiency
P deficiency occurs in acid uplands, calcareous soils, and acid sulfate
soils. It signifies the non availability of added or soil P to plants. Around
3 Mha of rainfed upland rice grown in humid zones of Africa and 6 Mha
in Latin America suffer from P deficiency. It is a problem in Ultisols,
Oxisols, Andosols, and some Vertisols.
Zinc (Zn) Deficiency
This is the most common nutrient deficiency of wetland rice after N and
P deficiency. It occurs in soils with pH more than 7 and organic carbon
more than 3%. Zinc deficiency in rice was first reported by Nene in 1966.
Subsequently more and more paddy lands suffering from it have been
identified. In India alone, it is suspected that around 8 Mha rice is
affected (Jones et al, 1982). It is common in calcareous, sodic, inland
saline, sandy, peat, and regardless of pH in continuously wet soils. Zinc
deficiency is becoming a problem in acid upland soils under reduced
fallow period in southeastern Nigeria (Singh et al, 1997). In the Ivory
Coast, it has also been observed in upland rice of the moist savanna zone
(Sahrawat et al, 1993).

B.N. Singh 223
Iron (Fe) Deficiency
Upland rice grown on acid, alluvial calcareous, alkaline and sandy soils
under anaerobic conditions suffers from iron deficiency. Iron is
important for chlorophyll biosynthesis and linked with a number of
enzyme systems in the plant. In India, it has been reported from Bihar,
West Bengal, and Vertisols of central and western Maharastra (Tandon
and Shinde, 1993). Crops such as soybean, chickpea, peanut, sugarcane,
and lentil also suffer from iron chlorosis.
Silicon (Si) Deficiency
Soils such as Oxisols, Ultisols, Histosols, and sandy Entisois are low in
available silica. Rice grown in acid upland soils of humid zones in Africa
and highly weathered upland soils in the savanna zone of Latin America
shows silicon deficiency. Use of basic slag as a source of Si to lowland
rice is common in Japan, Korea, and Taiwan for soils containing
relatively low levels of extractable Si.
Sulfur (S) Deficiency
Lack of organic sulfur causes stunted growth and yellowing of leaves. In
alkaline soils of Pakistan, soils having 33 ppm sulfur have shown
response to increased grain yield through sulfur application (Karim and
Majlish, 1958). Most of the wetland soils in India have started showing S
deficiency due to declining use of single superphosphate and
ammonium sulfate, and the sole application of urea as an inorganic
fertilizer. S deficiency not only affects plant yield, but also protein
quality, by reducing s3mthesis of S-containing amino acids.
Iron (Fe) Toxicity
This is one of the major soil constraints of acid lowland soils, inland
valley swamps, coastal swamps, and irrigated lowlands in Ultisols ¡and
Oxisols; ^In the humid forest and moist savanna zone of Africa, interflow
of ferrous ions occurs from upper slopes (Moormann and van Breeman,
1978). More than 50% lowland rice in Sierra Leone, Liberia, Guinea,
Nigeria, The Ivory Coast, and Senegal is affected by iron toxicity. It also
occurs in Sri Lanka, Vietnam, Malaysia, India (Kerala, Orissa), Indonesia
(Kalimantan and Sumatra), the Philippines, Brazil, Colombia, and
Madagascar (Sahrawat and Singh, 1995). Rice plants show symptoms of
broirzing due to high dissolved iron, Fe^'*' in soil solution around the

224 Rice Breeding and Genetics: Research Priorities and Challenges
rooting zone {Ponnamperuma et al, 1995). Young acid sulfate soils in
coastal areas also show symptorns of iron toxicity (Ottow et al, 1991).
Aluminum (Al) Toxicity
This is a problem in rice grown in acid uplands and acid sulfate soils^
with pH 5.0 and below. In such soils it is also related low P availability.
In acid uplands, if Al Concentration in soil solution is more than 1 to 2
ppm, it shows Al toxicity. The Al toxicity inhibits root growth and
restricts the nutrient and water uptake and leads to poor growth and
yield.
Hydrogen Sulfide (H2 S) Toxicity
H 2S toxicity occurs in young acid sulfate soils due to the reduction of
sulfates in submerged soils.
Boron (B) Toxicity
In coastal soils and volcanic areas, B toxicity is hazardous to crop
production. Soils irrigated with geothermal water or inundated by
brackish water have shown B toxicity.
Nutrient Imbalances
Nutrient stresses are caused by synergistic or antagonistic uptake of one
another. Deficiency or toxicity of one element may be induced by surplus
or toxicity of others. In iron toxic soils with a low leVel of P and K, the
toxicity symptoms are severe at low iron levels of 30 ppm. The ophmum
ratios for rice have been identified as: P /Z n (20-60), P /F e (10-20), K /
Na, Ca/M g (1.0-1.5), Fe/Z n (5-7), Pe/M n (1.5-2.5). More studies are
needed regarding these nutrient imbalances (De Datta et al, 1994).
1': ' j
VARIETAL IMPROVEMENT TO SOIL STRESSES
JRice is an ideal crop for reclamation of many problem soils. Selection of
rice cultivars with a higher level of tolerance to soil stresses has been an
ongoing research activity in many national and international programs.
Many superior cultivars were selected from the land races. International
agricultural research centers such as IRRI, CIAT, and WARDA have also
been engaged in the developments of improved cultivars with high
yield potential.

B.N. Singh 225
The production potential of rice cultivars in problem soils can be
enhanced through genetic manipulations- However^ there is need for a
better understanding of the constraints, their soil chemistry, nutrient
interactions development of reliable appropriate laboratory and field
screening methodologies, identification of hot spot locations and their
characterization, studies on the physiological mechanism of tolerance;
germplasm evaluation and breeding for tolerance, genetics of tolerance,
prebreeding and biotechnology; strengthening seed production, and
technology dissemination.
Diagnosis of Constraints
To develop appropriate varieties for problem soils, it is essential to
properly understand and characterize the soil constraint. Site
characterization is essential for better technology targeting. As the biotic
constraints vary from one location to another, it is relevant to develop
location-specific varieties. Sometimes tolerance to one stress is related to
other stresses. So it is better to evaluate for one stress initially and other
stresses later. Salinity tolerant lines are also tolerant to alkalinity, and P
and Zn deficiency. Fe toxicity tolerant lines are tolerant to Zn deficiency.
In rainfed uplands, tolerance to P deficiency is closely related to
tolerance to A1 or Mn toxicity. Thus, tolerance to salinity, Fe toxicity,
and P and Zn deficiencies are the crucial traits for developing stresstolerant
varieties.
Nutrient Interactions
It has been observed that in rice a deficiency of one nutrient causes
increased uptake of nutrients of the same valence. Under saline
conditions, K deficiency increased Na resulted in uptake (Yoshida and
Castaneda, 1969). Yoshida et at (1971) showed that Zn-deficient rice had
a higher Fe and Mn content than rice that had adequate Zn. N content is
also reduced in Zn-deficient rice (Sedberry et at, 1971). Iron toxicity is a
complex nutrient disorder and the deficiencies of other nutrients,
especially P, K, Ca, Mg, Si, and Zn, have been implicated for iron
toxicity in rice plants (Van Breemen and Moormann, 1978). A1 toxicity is
related to poor availability of P and Ca to the plants.
Symptoms and Screening Methodologies
Screening for different soil stresses can be carried out in the laboratory,
greenhouse and field. Screening methodologies for certain nutrients can
be developed in culture solutions and screening done in the greenhouse.

226 Rice Breeding and Genetics: Research Priorities and Challenges
Field screening at hot spot locations is also appropriate. But due to the
highly heterogeneous nature of the soil, several replications may be
needed for precision. There is a need to characterize the site for nutrients
in the root zone and also in the soil profile up to 100 or 150 cm.
Salinity tolerance screening in the laboratory is mostly confined to
the germination stage (Pearson et al, 1966; Bari et al, 1973). Yoshida et al
(1976) developed the greerütouse technique for salinity screening at
IRRI In this method, seeds are pregerminated on styrofoam with 100
holes and nylon net bottom. They are grown in a nutrient medium at pH
5.0 for 14 days. Further, they are transferred to a saline medium. The
salinity level of 6,800 ppm and EC 12 ds m"^ in solutiori is prepared by
adding 1:1 mixture of sodium chloride and calcium chloride. Seedling
survival is recorded after 2 weeks. Standard resistant and susceptible
checks are used for comparison (IRRI, 1996). Field screening at wellcharacterized
hot spot locations, or maintaining the pH and EC in the
field by irrigating with saline water can also be equally reliable. Asch et
al (1997) in a field screening method in arid sahel at Ndiaye, Senegal
maintained the sodium concentration (EC 3.5 mS cm“^) by irrigating
with sodium chloride saline water.
For A1 toxicity tolerance, both culture solution and field screenirig
techniques are available (Howeler and Cadavid, 1976; Sarkarung, 1986;
Coronel et al, 1990). In the culture solution method,.SO^ppm A1 is added
for the A1 toxicity screening. The relative root length, in 30. ppm A1 is
compared with that in the normal nutrient medium. In" field'screening,
comparison are made with reference varieties grown in?f trip plots
without lime and with 3 t lime plof^ (Sarkarung, 1986). ^ ;
Iron chlorosis is a problem of calcareous and sandy âéfhbic jsQilç. In
addition to field screening, a laboratory screening method By détiinating
orthophenanthroline reactive in the topmost leaves is simple; àffd
reliable (Singh et al, 1985, 1986).
In zinc deficiency, the symptoms start appearing a week after
transplanting and in the case of acute deficiency, the leaves dry and the
plant dies, In mild deficiency, sometimes plants recover, but mattirity is
delayed by 10-20 days. In early stages, the symptoms first appeal at the ^
third leaf from the top. Brown rusty spots appear toward the leaf base
that coalesce and grow toward the leaf tip. The zinc deficient plots haya
a brown rusty appearance, tillering is reduced, and growth is retarded.
IRRI (1996) has developed the Standard Evaluation System (SES) pf;
rice screening for salt and alkali injury, iron toxicity, and P and Zn
deficiencies. The SES is based on a scale of 1. to 9 :1 is highly .tolerant and
9 highly susceptible. Plant growth and tillering, are the two major traits
scored for screening. Symptoms of P deficiency are not easily
recognized. Tillering is severely reduced in P-deficiency. P-efficient

B.N. Singh 227
genot3^ e s have high tillering ability at low P levels in a culture solution
or in a P-deficient field. In salt injury/ the leaves roll up while in alkali
injury, they are discolored. In iron toxicity, the leaves turn orange,
orange-yellow, to reddish-brown or purple. Zinc deficiency symptoms
manifest at the third leaf from the top and the leaves turn brown.
Bronzing is the typical symptom of iron toxicity, but yellow or orange
coloration is observed in iron-toxic soils due to deficiencies in P, K, Ca,
and Mg induced by the high iron content (Howeler, 1973). It' is better to
evaluate the lines after flowering and at maturity. The comparison
should be made with tolerant and susceptible checks. As the soil is
highly heterogeneous, the maximum score among the different
replications should be taken as the criterion for reaction of a line. The
comparative yield in relation to tolerant check should also be used as a
criterion for selecting higher yielding genotypes. The scale for grain
yield will be different than the one for vegetative stage tolerance. Asch ef
al. (1997) classified lines with up to 40% yield reduction as tolerant, 41-
50% moderately tolerant, 51-60% moderately susceptible, and above
61% highly susceptible.
Hot-spot Locations
Certain sites, where the deficiency or toxicity symptoms always appear,
can be used for hot-spot screening (Table 10.1). In order to validate the
results from one site to other sites, and for technology targeting, it is
essential to characterize, the site for different nutrients. The growing
seasons also affect the manifestation of symptoms in the plants. In
salinity and iron toxicity screening under irrigated conditions at Ndiaye,
Senegal, and Korhogo Ivory Coast, higher damage scores were observed
in the dry season than in the wet season (Asch et al, 1997; Sahrawat and
Singh, 1998). So, where facilities are available, it is better to screen in the
dry season than in the wet season to select better tolerant lines.
Physiological Mechanism of Tolerance
It was earlier reported that both osmotic imbalance and an accumulation
of chloride ion () cause salt injury. Later the role of Na and Na-
K imbalance was reported as adversely affecting yield (Devitt et al,
1980; Ponnamperuma, 1984). Studies have shown that rice is very
tolerant to salinity during germination, but sensitive at the first to
second leaf stage, and at flowering. Its tolerance increases during
tillering (Pearson et al, 1966). Salinity damage is predominantly due to
excessive Na ion uptake and Na accumulation in the leaves. Resistance
to salinity is composed of avoidance and tolerance. Avoidance can be

228 Rice Breeding and Genetics: Research Priorities and Challenges
Table 10.1 Key sites for field screening for different problem soils
Soil stress
Inland salinity
Coastal salinity
Alkalinity and
sodicity
Iron toxicity
Aluminium
toxicity'
Zinc deficiency
Iron deficiency
Acid lowland
Acid uplands
Country and locations
Bangladesh (Joydebpur), Cambodia (Tuk Chha, Suay Rieng),
Burma (Kyauktan), Korea (Hwasong, Kyehwa, Namyang,
Gyhewa), Iran (Amol)
Bangladesh (Satkhira), India (Canning, -Panvel), Sri Lanka
(Ambalantota, Bentota, Kirinda), Egypt (Sakha, Seru), Sierra Leone
(Rokupr), Senegal (Djibelor)
India (Karnal, Kanpur, Kumarganj, Pindi), Thailand (Meung),
Pakistan (Pindi Bhattian)
Sierra Leone (Rokupr), Ivory Coast (Korhogo), Nigeria (Bende,
Edozhigi), Liberia (Suakoko), Guinea (Killissi)
Colombia (CIAT, llanos)
India (Pusa, Bihar), Philippines (IRRI)
India (Pusa, Bihar)
Thailand (Prachinburi), Burma (Myaung Mya, Sagamya), Malaysia
(Alor Setar), India (Barapani), Vietnam (Moc Hoa, Long an)
Burma (Aungban), Colombia (Villavicencio), Indonesia (Kotabaru),
Ivory Coast (Man), Nigeria (Amakama, Onne, Uyo), Philippines
(Iloilo), Vietnam (Dong Mai)
achieved by restricting the entry of sodium ions into the shoot (restricted
uptake, retention in the roots), whereas tolerance requires either
excretion through salt glands, or compartmentalization in stem, leaf
sheath, and older leaves (Flowers and Yeo, 1989). The potassiumsodium
absorption and its distribution in different plant parts is also an
important selection criterion for selecting salt-tolerant cultivars (Devitt
et ah, 1980; Gregorio and Senadhira, 1993; Asch et al, 1997). Tolerant
varieties such as Pokkali, Nona Bokra, and SR 26B excluded Na content
in shoots, absorbed more K, and had lower Na-K ratios (Gregorio and
Senadhira, 1993). Asch et al. (1997) studied Na-K ratio compartmentalization
in the top three leaves, older leaves, stem, and leaf sheath.
They observed lower ratios in tolerant cultivars in the top three leaves.
In response to sodium chloride, stressed plants accumulate
osmoprotective substances such as proline and trehalose. Seedling vigor,
vegetative growth, lodging, and plant height are also important traits
for dilution of salts, and should be considered for selecting salt-tolerant
rice varieties.
Zinc is an essential catalyst in the s)mthesis of auxins in plants. Zinc
deficiency reduces alcohol dehydrogenase (ADH) activity in roots under
anaerobic conditions after flooding and leads to a subsequent drop in
ATP production (Moore and Patrick, 1988).
Iron toxicity is mainly caused by an increase in ferrous concentrations
in the soil solution. Easily decomposable organic matter and
initial low soil pH reduce the ferric oxide (active Fe). Anaerobic
»!■

B.N. Singh 229
microbial activity also stimulates release (Ponnamperuma, 1965).
The critical level for iron toxicity in plants at maturity is 300 mg Fe kg“^.
The exact mechanism of varietal differences in tolerance to iron toxicity
is not clear, but it may be due to exclusion of iron in the oxidizing
rhizosphere, reduced translocation of iron, tolerance for high iron levels
in the plant tissues, or a combination of these factors (Jayawardena et ai,
1977).
Sufficient amounts of silica in rice plants enhance resistance to blast
and other fungal and bacterial pathogens associated with grain
discoloration (Winslow et al., 1997). It provides a glass like coating on
the epidermal surface that blocks penetration by pathogens.
Germplasm Evaluation and Breeding for Tolerance
Selection from the land race collections from a salt-affected area has
been one of the appropriate breeding methods (Table 10.2). Some
tolerant rice varieties Damodar, Dasal, Getu, Patnai 23, Pokalli,
Nonasail—are the selections from the land races of coastal saline areas
of West Bengal and Kerela, India (Bandhopadhyay and Sinha, 1985).
Jhona 349 was selected from the Punjab and Haryana for tolerance to
inland salinity. One of the early breeding lines of IRRI, IR 6-156-2, was
found suitable for saline soils of Sind province in Pakistan and is still
widely grown (Somoro and McLean, 1972). Mahsuri, a variety
developed in Malaysia, through the FAO Indica/Japonica hybridization
program, is P efficient and tolerant to zinc deficiency. It is the most
widely grown variety in India, Nepal, Bangladesh, and Myanmar. In
India, in the coastal saline area of Maharastra, Panvel varieties, and for
inland salinity rice varieties such as Usar 1, Co 43, PVR 1, MR 18, and
many other improved varieties have been developed by the pedigree
method (Salvi and Chavan, 1983; Rana, 1986), CSRIO, a high-yielding
semidwarf cultivar has been released for cultivation in inland saline
areas of India. It yields an average of 4 t ha"^ in salt-affected soils,
without amendments. Its cultivation for three consecutive years
improves soil and reduces sodicity stress (Mishra et al., 1992). IR 42 and
IR 64 are other improved semidwarf rice cultivars tolerant to P and Zn
deficiency, saliiuty, alkalinity, and iron and boron toxicity (Khush, 1987).
Both varieties are widely grown in many countries. From traditional
donors for salt tolerance, improved lines such as IR 4595-4-1-1-3
(Pokkali), IR 9884-54-3 and IR 10198-66-2 (Nona Bokra), and IR 10206-
29-2-1 (SR 26B) have been developed by the pedigree breeding method
(Chaubey and Senadhira, 1994), In coastal saline areas, where
waterlogging is a problem, intermediate stature, photosensitive varieties
are required. Through somaclonal variation, semidwarf lines from

230 Rice Breeding and Genetics: Research Priorities and Challenges
Pokkali were developed, which could be further used in breeding
programs (Senadhira et al, 1994).
Table 10.2
Traditional and inxproved rice varieties with a higher level of
tolerance to soil stresses
!•!■
Stress Traditional cultivars Improved cultivars
Inland Damodar, Getu, Dasal, Jhona 349, IR30,IR32,IR36,CSR10,
salinity Kalarata Citanduy, Usarl, Co43, MR18
Coastal Pokali, Rajasal, Nona Bokra, Nona sai IR8, Rok5, IR64, IR4630-22-2,
salinity 1, Patnai 23, Vy tilla 1, SR 26B Panvel 1
Akalinity
Acid
sulfate
Cheriviruppu, Damodar, Basmati
370, DA 29
KDML105, Bahagia
ÍR6(IR6-15ó-2)
SungaiLilin (IRl 1288-B-B-69-
1), NN2B (IR2823-399-5-6),
NN5B (IR4570-83-3-3), IR2151-
196-3-1-3>
IR 2153-26-3-5-6
Peat soils Bengawan, Kuatik Putih, Layang IR64, Sungallin (IR11288-B-B-
69-1)
P deficiency Patnai 23, SR 26 B, Jhona 349,
KDML 105, H4
IR5, IR20, IR28, IR42, IR54,
IR60, IR62, IR64, Mahsuri
Zinc Getu, Madhukar, Nam Sagui 19, BG90-2, IR20, IR32, IR34, IR42,
deficiency Pokkali, Bhura ratta Rasi, Govind (lET 6155), RAU
4009-3, RAU4005-26
Iron Azucena, Palawan, TCA148-3, TCA ÍR36,IR43,Rasi,MTU17,
deficiency 62-31-1 Prabhavati, IET7972, IET7973,
BR34
Iron toxicity
Al toxicity,
uplands
Al toxicity,
lowlands
Acid
uplands
Matcandu, Kuatik putih, Ngoba,
GÍSSÍ27
CR 94-13, ÍR20, Lemaya,
Suakoko 8,
WITAl, W rr A3, CK4, CK73
Azucena, Palawan, Moroberekan Ml-48 IRAT104, lAC 165,
IRAT 122
Salumpikit, Siyam Kuning,
IR29
Gudabang Putih, Siyam, Lemo
Azucena, M55, LAC 23, OS 6 HA 116, UPLRi-5, IRAT 144,
IRAT 104,IRATI33
Salinity tolerance is also a desirable trait in japónica rices. In the
screening of 657 japónica varieties, 19 lines were found tolerant.
However, none of the lines had tolerance similar to Pokkali. Namyang 7
was the best line (Lee and Senadhira, 1996). A field screening of 41
genotypes in strip plot design with fresh water and saline water
treatment was carried out at Ndiaye, Senegal during both wet and dry
seasons (Asch et ah, 1997). Grain yield decline, spikelet sterility, sodium
and potassium distribution in the top three leaves, stems, stem base and
roots, were taken as criteria for selecting salt tolerance. Five lines, I Kong
Pao, IR 64, IR 4630-22-2, CSR 10, and Aiwu were observed to be salt
tolerant. The potassium/sodium ratio in the three leaves from the top
was high in these genotypes and this could be used as a reliable index
for salt tolerance.

n
B.N. Singh 231-
Khao Dawk Mali 105, a selection from the land races, is a widely
grown variety in add sulfate soils of northeast Thailand. An IRRI line, IR
11288-B-B-69-1, has been released as Sungai Lilin in southern Sumatra,
Indonesia for acid sulfate soils. IR64, an early maturing semidwarf
variety, has been released as OM8 6 in Vietnam. It is also tolerant to add
sulfate soils. In peat soil at Kalimantan, Indonesia, some farmers have
started growing two crops in a year. High-yielding semidwarf cultivars
are grown as the first crop, followed by a local photosensitive crop. The
Surjan system has also been introduced in some areas (Ismunadji et al,
1991).
Genetic variability exists for tolerance to iron deficiency. Fe-efficient
genotypes can be screened at hot-spot locations. The problem of iron
chlorosis is severe in seedlings grown in dry seeded upland aerobic
soils. Certain genotypes and varieties have shown tolerance to iron
chlorosis. These are prabhavati, Rasi, MTU17, IR1561-22-8-3, Basmati 1-
63, PVRl, AUl (Tandon and Shinde, 1993). At Pusa, Bihar, India, the
problem is severe in calcareous saline sodic, soils. In a screening trial
involving 20 elite lines, only three lines, viz. BR 34, lET 7972, and lET
7973, were found tolerant (Singh et al, 1985, 1986).
Zinc deficiency is widespread in different parts of the Indian subcontinent.
In young and old alluvial calcareous soil of northern Bihar, 50
to 80% soils are zinc deficient. A field screening of 50 promising lines
was carried out at Pusa, Bihar (Singh et al, 1981). Based on the number
of hills infested in a 10 to 15 m^plot, and maximum score from 3
replications, only two lines, viz. RAU4005t26 and RAU4009-3, were
observed as tolerant and the other four, RAU4005-57, Govind (UPR 82-
1-7), RAU9-31-2-1, and IR2071-586-5-6-3 were moderately tolerant.
Other released varieties such as like Jaya, Prasad, Sita, Ratna, Rajendra
Dhanl, IR36 were found susceptible to highly susceptible. This is one of
the major reasons for low acceptability of modern semidwarf varieties
in northern Bihar.
Genetic evaluation and breeding for problem soil tolerance have
been systematically carried out at IRRI, Philippines. The selected donors
were used in hybridization programs and breeding lines were screened
for various traits like salinity, alkalinity, iron toxicity, peatiness, P, Zn
and Fe deficiencies, arid Al, Mn and B toxicities; Up to December 1992,
about 200,000 lines had been screened at IRRI, and about 15% were
tolerant (De Datta et al, 1994), Many improved lines with higher yield
potential and resistance to different diseases and pests have been
developed from traditional donors. Some of them are already being
grown in many problem soil areas (Table 10.2) Rapid generation advance
(RGA) and shuttle breeding were used to raise two to three generations
each year. Anther culture of Fj allows the rapid fixation of homozygosity

232 Rice Breeding and Genetics: Research Priorities and Challenges
1.
and significantly reduces the breeding cycle, Zapata et al (1991)
transferred the salinity tolerance from tolerant varieties through anther
culture. One of the lines ACj, performed well under botii saline and non
saline conditions.
Screening for iron toxicity was one of the major breeding objectives
at WARDA research station at Suakoko, Liberia, UTA and WARDA
research stations at Edozhigi, Nigeria, and Korhogo, Ivory Coast. For
iron toxicity, over 30,000 lines have been screened so far at different
stations of WARDA and IITA. Most of the lines, including such highyielding
varieties as IR5, Bouake 189, ITA212, ITA306, BG90-2, IR46,
ITA123, were severely affected by bronzing and were susceptible,
Matcandu and Gissi 27 were identified as donors for tolerance to iron
toxicity. Suakoko 8 , and Suakoko 1 2 were released as varieties tolerant
to iron toxicity from WARDA center in Liberia. Suakoko 8 was later
released in Sierra Leone as ROK 24 and is still a widely grown variety of
inland valley swamps, Suakoko 8 was selected from a cross between
Siam 25 and Malunja and introduced from Malaysia as line 2526 in 1973
(Virmani et al, 1978), WITA 1 and WITA 3 were recently released in the
Ivory coast for iron toxicity areas. CK4, CK73, and CK263 are the other
varieties released in Guinea Conakry. WITA 1 and WITA 4 have been
found tolerant to iron toxicity in Nigeria (Sahrawat and Singh, 1995,
1998). In Indonesia, varieties such as Kapuas and Batang Ombilin, and
in Sri Lanka-BWlOO, BW267-3, BW272-6B, are recommended for iron
toxic areas (Ismumadji et al, 1991; Gunatilaka, 1994),
For Al toxicity and P deficiency, screening methods in culture
solution have been developed at IRRI to screen a large number of lines
(Chaubey et al, 1994, Khatiwada et al, 1996). Tolerance to Al toxicity is a
major breeding objective at CIAT Cali, Colombia. A large number of
lines are screened routinely (Sarkarung, 1986), Only tolerant lines are
further evaluated in yield tiials. Screening for tolerance to acid uplands
is one of the major breeding objectives at WARDA, Bouake, Ivory Coast.
Screening of lines is routinely carried out near Man. The soil is an Ultisol
and tolerance to P deficiency is one of the major selection criteria. Some
of the newly released varieties, e.g. WAB56-50, WAB56-125, and
WAB56-104, had better P efficiency than IDSA6 (Sahrawat et al, 1997).
Tropical japónica varieties were more adapted to add upland
ecology in West Africa and Latin America. Winslow et al, (1997)
observed that they have higher Si content (93%) in their hush compared
to indica genotypes. They also exhibit better resistance to grain
discoloration than indicas. These differences between the ecotypes
suggest their adaptation mechanism to Si-deficient soils.

?
B.N. Singh 233
GENETICS OF TOLÉRANCE
Genetic studies of tolerance have been carried out for certain traits such
as salt tolerance, iron toxicity, P efficiency, and A1 toxicity tolerance.
Jones (1986) studied the genetics of salt tolerance in two mangrove
swamp varieties, viz. Pokali and Pa Merr 108A at Rokupr, Sierra Leone.
The relative root length in saline and nonsaline culture solution was
taken as a criterion for measuring salt tolerance. Additive genetic
variation, maternal effects, and transgressive segregation were observed
in parents and progenies. Akbar et al. (1986) studied the genetics of salt
tolerance in Pokkali, Nona Bokra, Damodar and Jhona 349. Growing
plants in a nutrient solution medium, Yoshida et al. (1976) carried out
genetic analysis. Data were taken on Na and Ca levels at the seedling
stage, and yield per plant. Both additive and dominance effects were
observed for the traits studied.
Gregorio and Senadhira (1993) reported the genetics of salinity
tolerance in Nona Bokra, Pokkali, and SR 26B using the culture solution
method. Na-K ratio was used as a criterion for selecting parents and
crosses with high general and specific combining abilities. Improved
lines were good general combiners and reciprocal differences were
observed, suggesting that tolerant parents should be used as female
parents in crosses.
Abifarin (1986) reported a single dominant gene in Suakoko 8 and a
recessive gene in Gissi 27 for tolerance to iron toxicity. The genes in two
resistant lines were nonallelic, which shows that more tolerant plants
could be recovered from the segregating generations of the two crosses.
Chaubey et al. (1994) studied the genetics of P deficiency in lowland rice
varieties. Relative ability to tiller under P-deficient and P-suppIemented
nutrient medium was used to classify the genotypes. The tolerant
parents—^IR 20, IR 28, IR 54 and Mahsuri— showed recessive genes for
P efficiency. Both additive and dominant gene effects were observed in
the tolerant parents. IR 54, Mahsuri and IR20 were good general
combiners. Khatiwada et al. (1996) have studied the genetics of Al
toxicity in upland and lowland rice cultivars, IRAT 104, Moroberekan,
IR 43 and IR 29. In a diallel study of parents and F|, upland varieties
were found to be good general combiners. Reciprocal effects were also
observed. In such situations, tolerant parents should be used as female
parents in crosses.
PREBREEDING AND BIOTECHNOLOGY
In coastal areas of India and Bangladesh, plants of Porteresia coarctata, of
the tribe oryzeae, contain high Na and tolerate high salinity. The

234 Rice Breeding, and Genetics: Research Priorities and Challenges
mechanism seems to be tissue tolerance. Oryza rufipogon and O.
glaberrima germplasm are known for their tolerance to acidity and iron
toxicity. There is a need to select donors with a high degree of tolerance
and to incorporate these traits into improved germplasm. Somaclonal
variation has been used to select high-yielding salt tolerant lines, which
can also be used as donors in breeding programs (Senadhira et al, 1994).
In bread wheat, Triticum aestimm L., a single locus, Knal, has been
identified, which controls Na"^ exclusion in roots and enhances K^/Na"^
ratio in the shoots and salt tolerance (Dvorak et at, 1994). The existence
of such genes in rice should be explored. Recent biotechnological tools
can help in isolating such genes, cloning and incorporating them in a
high-yielding background, even from the related and alien species.
GERMPLASM EXCHANGE, SEED PRODUCTION AND
TECHNOLOGY DISSEMINATION
Tolerance to nutrient toxicity or deficiency is one of the major traits for
selecting higher yielding cultivars for problem soils. Growth duration,
non-lodging, resistance to location-specific disease and insect resistance,
photoperiod sensitivity, grain type, pericarp color and other attributes
decide the acceptability of a line. In temperate regions, tolerance to cold,
and grain quality are some other major criteria for varietal selections.
The earliest recorded introduction and germplasm exchange of a salttolerant
variety is Pokkali, a selection from a land race in Kerala, India,
to Sri Lanka in 1939. It was recommended for cultivation in 1945 on
saline rice lands of the West Coast (Fernando, 1949). IRRI, through its
International Rice Testing Program (IRTP) since 1976 and later through
the International Network for Genetic Evaluation in Rice (INGER), has
played a major role in germplasm exchange among different ricegrowing
countries. The two nurseries viz., IRSATON (International Rice
Salinity and Alkalinity Tolerance Observational Nursery), and IRALON
(International Rice Acid Lowland Soils Observational Nursery) were
involved in distribution of lines from ÎRRI and national programs. In
West Africa, improved germplasm and segregating populations are
distributed to NARS through the lowland rice breeding task force
(Sahrawat and Singh, 1995).
INTEGRATED MANAGEMENT
A tolerant variety alone cannot increase the production potential from
adverse soils. But growing of a tolerant variety will reduce the dosage of
plant nutrients and soil amendments. It is essential to take an integrated
approach to increase productivity from such soils. Green manuring with

B.N. Singh 235
I
Sesbania is a useful practice for rnanagement of saline and sodic soils.
Rice is the most favored crop in such soils. Nitrogen up to 160 kg in 3-^
splits as ammonium sulfate has been found better than urea^ and can
yield up to 8 t ha'\ Use of slow release fertilizers and sulfur-coated or
lac-coated urea has shown promise in saline soils. Use of gypsum and
leaching reduces sodidty. For management of iron-toxic soils^r cultural
practices such as early plantings drainage of fields and planting on the
ridges (Winslow et ai, 1989), use of sufficient P, K, Ca, Mg, Si, and Zn
nutrients, in addition to planting tolerant varieties, will increase the
production potential from such soils (Fatra and Mohanty, 1989;
Sahrawat and Singh, 1995).
Application of P and K should be increased one and half times over
the normal dose. Use of basic slag at the rate of 10--151 ha“^will reduce
the silicon deficiency in soils. Either soil or foliar application with zinc
sulfate can correct Zn deficiency. Application of zinc sulfate (20% zinc)
at the rate of 25 kg ha'^ suffices for growing six crops. When symptoms
of zinc deficiency appear in the field, a spray of 1 % zinc sulfate with
0.5% lime will alleviate the deficiency symptoms. Application of rock
phosphate has been suggested in acid upland soils (Panda, 1987). In acid
sulfate soils, liming, P and K application, and drainage where possible
reduce the adverse effect in such soils. In the Mekong delta, 3-6 tons of
lime, 22 kg P ha'^, and an early plowing immediately after flood
recession is recommended to manage the acid sulfate soils with high Al
toxicity (Van Mensvoort et al, 1985). Peat soils can be productive by
growing wetland rice with tolerant varieties, and N, P, K, and Zn
fertilizer use. Sulfur deficiency can be corrected by application of sulfur
at the rate of 11 kg ha'^. Application of pyrites, gypsum, ammonium
sulfate, manganese sulfate, and single superphosphate alleviates S
deficiency. Raising paddy seedling under puddled or flooded condition
will reduce iron chlorosis. Soil application of ferrous sulfate has not
proven to be a corrective measure. Two to three foliar sprays of 1%
ferrous sulfate or Fe-chelates, mixed with 1% lemon (vitamin C) will
reduce chlorosis damage (Tandon and Shinde, 1993). Application of
lime and P fertilizer alleviates Al toxicity damage in acid upland. Seawater
intrusion in acid sulfate soils reduces the effect of Al toxicity.
CONCLUSIONS AND FOLLOW-UP
Problem soils are the most potential areas of food production to meet
the growing demand of the increasing world population. Productivity
from such soils can be increased through selecting tolerant varieties and
applying appropriate soil amendments. Characterization of problem

236 Rice Breeding and Genetics: Research Priorities and Challenges
soils is the primary step to develop an integrated approach for
management of such soils. Screening for salinity, zinc deficiency, P
defficiency, and iron toxicity should be the four major areas for largescale
screening. Prebreeding will help in broadening the genetic and
physiological basis of resistance. Biotechnology can help in tagging,
cloning, and incorporating genes for tolerance to different stresses.
There is a need to share the available and new information through
small working groups. Farmer participation in problem identification,
technology testing and transfer, technology targeting, and training are
essential. Modeling would provide an essential tool in characterization
and problem solving. Seed production of new varieties and availability
of soil amendments at the proper time are essential in technology
dissemination. International centers such as IRRI, CIAT, WARDA, and
IWMI in collaboration with institutions in developed countries, NARS,
and NGOs should strengthen the upward and downward research.
There is need to have task forces to provide small grants for thematic
and focussed research and development activities. Public awareness
and annual mettings on "Management of problem soils for rice
cultivation" at different centers should be essential activities in future.
Acknowledgment
The author expresses his gratitude to Dr. K.L. Sahrawat, soil scientist at
WARDA, for valuable discussion and critical comments on the
manuscript.
References
i!« >:
Abifarin, A.O. 1986. Inheritance of tolerance to iron toxicity in two rice cultivars. In: Rice
Genetics, IRRI, Manila, Philippines, pp. 423-428.
Abrol, I.P. 1986. Salt affected soils: Problems and prospects in developing countries. In:
Global Aspects of Food Production M.S. Swaminathan and S.K. Sinha (eds.) IRRI, Manila,
Philippines, pp. 283-305.
Akbar, M. 1975. Water and chloride absorption in rice seedlings. /. Agrie. Res. 13:341-343.
Akbar, M., Khush, G.S. and HilleRis Lambers, D. 1986. Genetics of salt tolerance in rice. In:
Rice Genetics, pp. 399-409.
Anjos Cutrim, V.D., Ngugen, T. V., Silva, f.C. and Galvo, J.D. 1981. Inheritance of tolerance to
aluminium toxicity in Brazilian rice. Int. Rice Res. Newslett. 6(4): 9.
Asch, F., Dingkuhn, M. and Dorffling, K. 1997. Physiological stresses of irrigated rice caused
by soil salinity in the Sahel. In: Irrigated Rice in the Sahel: Prospects for Sustainable Development
K.M, Miezan, M.C.S. Woperies, M,Dingkuhn, J, Deckers, T.F. Randolph (eds.).
West Africa Rice Development Association, pp. 247-274.
Bandyopadhyay, A.K. and Sinha, T.S. 1985. Rice Cultivation in Coastal Saline Soils. Better
Fanning in Salt Affected Soils. Series 10. Central Soil Salinity Res. Inst., Karnal, India.
Bari, G., Hamid, A. and Awan, M.A. 1973. Effect of salinity on germination and seedling
growth of rice varieties. IRC Neivletl. 22 (3): 32-36.

I j
B.N. Singh 237
Boje-Klein, G. 1986. Problem soils as potential areas for adverse soils tolerant varieties in
south and southeast Asia. IRRI Research Paper Series Vol. 119,53 pp.
Chaubey, C.N. and Senadhira, D. 1994. Conventional plant breeding for tolerance to problem
soils. Theor. Appl. Genet. Monograph, 21:11-36.
Chaubey, C.N., Senadhira, D. and Gregorio, G.B. 1994. Genetic analysis of tolerance for
phosphorus deficiency in rice. Theor. Appl. Genet. 89: 313-317.
Coronel, V.P., Akita, S. and Yoshida, S. 1990. Aluminium toxicity tolerance in rice seedlings,
la Plant Nutrition-Physiology and Applications M.L. Van Beusichem (ed.). Kluwer
Academic Publ., The Netherlands pp. 357-363.
De Datta, S.K., Neue, H.U., Senadhira, D. and Quijano, C. 1994. Success in rice improvements
for poor soils 1994. In: Proc. Workshop on Adaptation of Plants to Soil Stresses. August 1-4,
1993, Univ. Nebraska, Lincoln, Nebraska.
Devitt, D., Jarrell, W.M., Stevenes, K.L. 1980. Sodium-Potassium ratios in soil solutions and
plant response under saline conditions. Soil Sei. Soc. Am. J. 45:80-86.
Driessen, P.M, 1978. Peat Soils, In: Soils and rice, IRRI, Manila, Philippines, pp. 763-779.
Dvorak, J., Noaman, M.M., Goyal, S. and Gorham, J. 1994. Enhancement of the salt tolerance
of Tniicwm turgidum by the Knal locus transferred from theTnfici/m aestivum chromosome
4D by homeologous recombination. Theor. Appl. Genet. 87:872-877.
Fernando, L.H. 1949. The performance of salt resistant paddy, Pokkali, in Ceylon. Trop.
Agric. 105:124-126.
Flowers, T J. and Yeo, A.R. 1989. Effects of salinity on plant growth and crop yield. In:
Environmental Stress in Plantsj.H. Cherry (ed.). Springer-Verlag, Berlin, Heidelberg.
Flowers, T.J. and Yeo, A.R. 1995. Breeding for salinity resistance in crop plants: where next?
Aust. J, Plant Physiol. 22; 875-884.
Ghassemi, F., Jakeman, A.J. and Nix, H.A. 1995. Salinisation of land and water resources.
Human causes, management and case studies. Canberra, Australia. Centre for Resource
and Environmental studies.
Gleick, P.H, 1993. Water in crisis; A guide to the world's fresh water resources. Oxford
University Press, New York, USA.
Gregorio, G.B. and Senadhira, D. 1993. Genetic analysis of salinity tolerance in rice. Theor.
Appl. Genet. 86: 333-338.
Giinatilaka, G. A. 1994. Rice production in relation to iron toxic soils in Sri Lanka, In: Rice and
Problem Soils in South and Southeast Asia. D. Senadhira (ed.). IRRI Discussion paper series,
4, IRRI, Manila, Philippines pp. 109-114.
Gupta, R.J. 1994. Groundwater quality for irrigation. In: Salinity Management for Sustainable
Agriculture. D.L.N. Rao, N.T. Singh, R.K. Gupta, and N.K. Tyagi. (eds.) Central Soil
salinity Research Institute, Karnal, India, pp. 89-100.
Howeler, R.H, 1973. Iron induced oranging disease of rice in relation to physiochemical
changes in a flooded oxisol. Soil Sei. Soc. Amer. Proc. 37:898-903.
Howeler, R.H. and Cadavid, L.F. 1976. Screening for rice cultivars to A1 toxicity in nutrient
solution as compared with a field screening method. Agron.}. 68:551-555.
Huang, J. and Rozelle, S. 1993. Environmental stress and grain yield in China. Resource
paper for the Dubos forum. Technology Prospects for Sustainable Agriculture.
Ikehashi, H. and Ponnamperuma, F.N. 1976. Varietal tolerance of rice for adverse soils. In:
Soils and Rice. IRRI, Manila, Philippines, pp 801-823.
IRRI. 1996. Standard Evaluation System for rice (4th ed.), July 1996, INGER, IRRI, Manila,
Philippines, 52 pp.
Ismunadji, M. and Soepardi, G. 1984. Peat soils problems and crop production. In: Orgawic
Matter and Rice. IRRI, Manila, Philippines, pp. 489-502.

238 Rice Breeding and Genetics: Research Priorities and Challenges
iim i
i lii ;■
■ i
I
Ismunadji, M., Ardjasa, W.S, and Von Uexkull, H.R. 1991. Increasing productivity of iron
toxic soils of Indonesia. In: Rice Production on the Acid Soils of the Tropics, pp. 213-217. P,
Deturck and F.N. Ponnamperuma (eds,). Institute of Fundamental Studies, Kandy,
Sri Lanka.
Jayawardena, S.D.G., Watabe, T. and Tanaka, K. 1977. Relation between oxidizing power
and resistance to iron toxicity in rice. Rep, Soc. Crop Sei. Breed. Kinki, Japan 22:35-47.
Jones, M.P. 1986. Genetic analysis of salt tolerance in mangrove swamp rice. In: Rice Genetics.
IRRI, Manila, Philippines, pp. 411-422.
Jones, U.S., Katyal, J.C., Mamaril, C.P. and Park, C.S. 1982, Wetland rice nutrient deficiencies
other than Nitrogen. In; Rice Research Strategies for the Future. IRRI, Manila, Philippines
pp. 327-378.
Karim, M. and Majlish, M.A.K. 1958. A study of the formative effects of sulphur on rice
plant. Pak. J. Sei. Res. 10:52
Khatiwada, S.P., Senadhira, D., Carpena, A.L., Zeigler, R.S. and Fernandez, P.G. 1996.
Variability and genetics of tolerance for aluminium toxicity in rice. Theor, Appl. Genet. 93:
738-744.
Khush, G.S. 1987. Rice breeding: past, present and future. J. Genet. 66(3): 195-216.
Kijne, J.W., Prathapar, S.A., Wopereis, M.C.S. and Sahrawat, K.L. 1998. How to manage
salinity in irrigated lands: A selective review with particular reference to developing
countries. SWIM Paper 2, International Irrigation Water Management Institute, Colombo,
Sri Lanka.
Lee, K.S. and Senadhira, D. 1996. Salinity tolerance in Japónica rice. SABRAO /. 28 (1); 11-17.
Maas, E.V. and Hoffman, G.J. 1997. Crop salt tolerance: current assessment./. Irrg. Drainage
Div. ASCE 103:115-134,
Mishra, B., Singh, R.K. and Bhattacharya, R.K. 1992, CSR10, a newly released dwarf rice for
salt affected soils. Int. Rice Res. hiotest 17(1); 19.
Moore, P.A. and Patrick, W.FL 1988. Effect of zinc deficiency on alcohol dehydrogenase
activity and nutrient uptake in rice. Agron.}. 80 :882-885.
Moorman, P.R. and Van Breeman, N. 1978. Rice: Soil, Water, and Land. IRRI, Manila,
Philippines, 185 pp.
Munns, R. and Richards, R.A. 1998. Improving crop productivity in saline soils. In: Crop
Productivity and Sustainability: Shaping the Future Proc. 2nd Int. Crop Sei. Cong. 1996. L.
Chopra, R.B, Singh and A. Verma (eds.), Oxford and IBH Publishing, New Delhi, pp. 453-
465.
Nene, Y.L. 1966. Symptoms cause and control of Khaira disease of paddy. Bull. Indian
Phytopath. Soc. 3; 97-101.
Ottow, J.C.G., Prade, K., Bertembreiter, W, and Jacq, V.A. 1991. Strategies to alleviate iron
toxicity of wetland rice on acid sulphate soil. In: Rice Production on Acid Soils of the tropics.
P. Detruck and Ponnamperuma (eds.). Institute of Fundamental Studies, Kandy, Sri
Lanka.
Panda, N. 1987. Acid soils of eastern India: Their chemistry and management. /. Indian Soc.
Soil Sn. 35:568-581.
Patra, B.N. and Mohanty, S.K. 1989. Effects of amendments on transformations of Fe and Mn
in iron toxic soils under submergence. Ind. Soc. Soil Sei, 37:276-283.
Pearson, G.A., Ayers, A.D. and Eberhard, D.L. 1966. Relative salt tolerance of rice during
germination and early seedling development. Soil Sei. 102:151-156.
Ponnamperuma, F.N. 1965, Dynamic aspects of flooded soils and the nutrition of rice plant.
In; The Mineral Nutrition of the Rice Plant. John Hopkins Press, Baltimore, Maryland,
pp. 295-298.

vt
B.N, Singh 239
Ponnampemma, F.N. 1984. Role of cultivar tolerance in increasing rice production in saline
lands. In; Salinity Tolerance in Plants. Strategies for Crop Improvements. R.C. Staples and
G.H. Toenniessen. (eds.). Wiley Inlerscience, New York pp 255-271.
Ponnamperuma, F.N. and Bandyopadhya, A.K. 1980. Soil salinity as a constraint on food
production in the humid torpics. In: Priorities for Alleviating Soil Related Constraints to Food
Production in the Tropics. IRRh Manila^ Philippines, pp. 203-216.
Ponnamperuma, F.N., Bradfield, R. and Peek, M. 1955. Physiological disease of rice
attributable to iron toxicity. Nature 175:265,
Rana, R.S. 1986. Breeding crop varieties for salt affected soil. In: Approaches for Incorporating
Drought and Salinity Resistance in Crop PÍBnfs. V.L, Chopra and R.S.Paroda (eds.). Oxford
and IBH, New Delhi, pp 25-55.
Rosegrant, M.W., Agcao, li M. and Perez, N. 1995. Rice and Global Pood Economy:
Projections and Policy Implications of future Food Balances. International Food Policy
Research Institute, Washington, D.C.
Sahrawat, K.L. and Singh, B.N. 1995. Management of iron toxic soils for lowland rice
cultivation in West Africa. Proc. Third African Soil Science Society Conference, Vol. 28.
pp. 617-624.
Sahrawat, K.L. and Singh, B.N. 1998. Seasonal differences in iron toxicity tolerance of
lowland rice cultivars. Int. Rice Res. Notes: 23(1): 18-19,
Sahrawat, K.L., Diatta, S. and Jones, M.P. 1993. Zinc deficiency in upland rice. WARD A Ann.
Rep. 1993. pp. 37-38.
Sahrawat, K.L., Jones, M.P. and Diatta, S. 1997. Direct and residual fertilizer phosphorus
efficiency of upland rice in an ultisol. Nutrient Cycling in Agrocecosystems. 48; 209-215.
Sahrawat, K.L., Mulbah, C.K., Diatta, S., Delaune, R.D., Patrick, W.H., Singh, B.N. and Jones,
M.P. 1996. The role of tolerant genotypes and plant nutrients in the management of iron
toxicity in lowland rice. /. Agrie. Sei, Cambridge. 126:143-149.
Salvi, P.V. and Chavan, K.N. 1983. Coastal saline soils of Maharashtra. /. Indian Soc. Coastal
Agrie. Res. 1:21-26.
Sarkarung, S. 1986. Screening upland rice for aluminium tolerance and blast resistance, In:
Progress in Upland Rice Research. IRRI, Manila, Philippines, pp. 271-281.
Sedberry, J.E., Petreson, F.J., Wilson, E., Nugent, A.L., Engler, R.M, and Brupbacher, R.H.
1971. Effect of zinc and other elements on the yield of rice land nutrient content of the rice
plants. Louisiana State Univ. Agrie Expt. Stn. Bull. 653.
Senadhira, D., Neue, H.U. and Akbar, M. 1994. Development of improved donors for salinity
tolerance in rice through somaclonal variation. SABRAO J. 26:19-25.
Singh, B.N. and Singh, B.P. 1984. Tolerance to iron deficiency in rice. Int. Rice Res. Newslett.
9(6): 8.
Singh, B.N., Singh, B.P. and Sinha, M.K. 1981. Relative tolerance of semi-dwarf indica rice
lines to zinc stress in calcareous soils. Ann. Agrie. Res. 2(1-2): 27-32.
Singh, B.N., Fagade, S., Ukwungwu, M.N., Williams, C., Jagtap, S. S., Oladimeji, O., Efisue, A.
and Okhidievbie, O. 1977. Rice growing environments and biophysical constraints in
different agroecological zones of Nigeria. Mei. /. 2(1): 35-44'.
Singh, P.B., Singh, R.A., Sinha, M.K. and Singh, B.N. 1985. Evaluation of technique for
screening Fe-efficient genotypes of rice in calcareous soil. J. Agrie. Sei, Cambridge: 105:
193-197.
Singh, B.P., Sinha, M.K., Singh, R.A. and Singh, B.N. 1986. Reaction of genotypes of rice to
iron chlorosis in a calcareous soil. Exp. Agrie. 22:75-78.
Somoro, A.A. and McLean, G.W. 1972. High yielding rice varieties in West Pakistan. In; Ríce;
Breeding. IRRI, Manila, Philippines, pp. 157-159.

ii'
240 Rice Breeding and Genetics: Research Priorities and Challenges
Tagawa, T. and Ishizaka, N. 1963. Physiological sttidies on the tolerance of rice plants to
salinity. Proc. Crop Sci. Soc. Jpn. 31: 249-252.
Tandon, H.L.S. and Shinde, J.E. 1993. Fertility management for rice in problem soils of India.
In; Neiv Frontiers in Rice Research; K. Muralidharan and E.A. Siddiq. (eds.). Directorate of
Rice Research, Hyderabad, India, pp. 315-322,
Van Breeman, N. and Moormann, F.R. 1978. Iron Toxic soils. In: Soils and Rice. IRRI, Manila,
Philippines, pp. 781-800.
Van Mensvoort, M.E., Lantin, R.S., Brinkman, R. and Van Breeman, N. 1985. Toxidties of
wetland soils. In: Wetland. Soils: Characteristation, Classification, and Utilisation. IRRI
Manila, Philippines, pp. 123-138.
Virmani, S.S., Tubman, A.F., Sumo, F. and Worzi, P.M, 1978. Suakoko 8, a new rice variety
recommended for iron toxic swamps in Liberia. Int. Rice Res. Nezoslett, 3; 3~4,
Winslow, M.D., Okada, K. and Correa-Victoria, F. 1997. Silicon deficiency and the adaptation
of tropical rice ecotypes. Plant and Soil. 188: 239-248.
Winslow, M.D., Yamauchi, M., Alluri, K. and Masajo, K., 1989. Reducing iron toxicity in rice
with resistant genotype and ridge planting. Agron. J. 81:458-460.
Yoshida, S. and Castaneda, L. 1969. Partial replacement of potassium by sodium In the rice
plant under weakly saline condition. Soil Sci. Plant Nutr. 15; 183-186.
Yoshida, S., Porno, D.A. and Bhadrachalam, A. 1971. Zinc deficiency of the rice plant on
calcareous and neutral soils in the Philippines. Soil Sci. Plant Nutr. 17: 83-87.
Yoshida, S., Forno, D.A., Cock, S.H. and Gomez, K.A. 1976. Laboratory Manual for
Physiological Studies in Rice. IRRI, Manila, Philippines.
Zapata, F.J., Alejar, M.S., Torrizo, L.B., Novero, A.U., Singh, V.P. and Senadhira, D. 1991.
Field performance of anther culture derived lines from Fj crosses of indica rices under
saline and nonsaline conditions. Theor. Appl. Genet. 83:6-11.
i I
■> t

u
Molecular Marker-Based
Gene Tagging and Its Impact
on Rice Improvement
Qifa Zhang* and Sibin Yu*
INTRODUCTION
Developments of biotechnology in the last two decades have resulted in
two main technical approaches for genetic manipulation of the plant
genome, transformation, and marker assisted selection. While
transformation is dependent on the availability of genes that are isolated
and cloned and is the subject of a different chapter, the technique of
marker-assisted selection is completely based on mapping and genetic
characterization of the genes for the targeted traits. Recent advances in
genome research and the availability of high-density molecular marker
linkage maps (Causse et ah, 1994; Kurata et ah, 1994) have greatly
facilitated gene mapping and genetic analyses in rice. In this chapter, we
shall focus on the present status of gene mapping and genetic analyses of
the traits that are the major targets for rice-breeding programs worldwide.
We shall also discuss the impacts of such developments on rice
genetic improvement.
* National Key Laboratory of Crop Genetic Improvement, Huazhong Agricultural University,
Wuhan 430070, China

242 Rice Breeding and Genetics: Research Priorities and Challenges
MAPPING AND GENETIC ANALYSIS OF DISEASE RESISTANCE
i
i i'i'f
Monogenic Resistance
Disease resistance genes have been one of the major subjects for
molecular marker-based mapping and genetic analyses. A large number
of studies have been carried out in the last decade for identifying and
mapping genes for resistance to various rice diseases (Table 11.1). All
classes of available molecular markers have been used in the mapping
studies including restriction fragment length polymorphisms (RELPs),
randomly amplified polymorphic DNA (RAPDs), simple sequence
repeats (SSRs), sequence tagged sites (SSTs)^ and amplified fragment
length polymorphisms (AFLPs). Near isogenic lines (NILs) that were
developed by introgressing the resistance gene from the donor parent
via repeated back crossing were used extensively for identifying the
linked markers, hence the gene-containing chromosomal regions.
Segregating populations from crosses between NILs was also frequently
used to determine the linkage between molecular markers and the
targeted genes. When NILs were not available, bulked segregant
analysis was also used for identifying linked markers and segregating
populations, or in some cases only susceptible individuals were used for
calculating the recombination frequencies between markers and the
genes.
The studies listed in Table 11.1 were highly concentrated in the two
most important diseases, fungal blast caused by Pyricularia grísea and
bacterial leaf blight (BLB) caused by Xanthomonas oryzae pv. oryzae, that
collectively account for most of the yield losses of rice worldwide due to
diseases. Molecular markers linked to the genes at distances within a
few centi-Morgan (cM) were identified for the majority of the genes, and
markers flanking the resistance genes on both sides were also obtained
in a number of cases. In the extreme, cosegregating markers were obtained
for several genes. However, there were also cases in which the
closest markers were not tightly linked to the genes, especially for
markers identified for blast resistance genes. Additional markers therefore
should be identified for those genes. It should also be noted that
linked molecular markers have not been found for quite a few blast
resistance genes although the chromosomal locations were determined
previously (McCouch et ah, 1994).
Inspection of the genes listed in Table 11.1 clearly revealed an
important feature regarding their distribution. A large cluster of disease
resistance genes is located on chromosome 1 1 , which includes genes for
resistance to BLB, blast, and viruses. There also appeared to be another
cluster of genes on chromosome 12. The individuality of the resistance

Qifa Zhang and Sibin Yu 243
genes on chromosome 1 1 appears to be evident as judged by the mapping
results; each of the BLB resistance genes Xa3, Xa4, XalO, Xa21^ and
Xa22 is mapped to a distinct position. However^ the identities of the
blast resistance genes are less clear and need further characterization.
For example, the genes Pi-ta and Pi-ta^ could not be separated even in
large F2 populations (Rybka et al., 1997), It has also been suggested that
Pi-4 is allelic or possibly identical to Pi-ta (Inukai e t 1994).
Quantitative Resistance
In addition to the monogenic resistance that confers complete resistance
and is inherited qualitatively, many systems of host-pathogen
interactions often result in partial resistance or quantitative resistance
(Young, 1996). While qualitative resistance conditions compatibility
between the host carrying the specific resistance gene and pathogen
strain carrying the corresponding a virulence gene, quantitative
resistance reduces the level of disease damage in a compatible reaction.
Because quantitative resistance places less selection pressure on the
specific pathogen strain than does qualitative resistance, it is also
believed to be more durable than qualitative resistance.
There have been several undertakings in mapping quantitative resistance
in rice. An outstanding example is the work of Wang et al. (1994)
who conducted a quantitative trait locus (QTL) analysis of the blast
resistance in the variety "Morobereken" in! various environments. The
experimental population consisted of 281 F7 recombinant inbred lines
(Rlt-s) from a cross between two varieties, '"Morobereken^'— a japónica
cultivar with durable resistance to blast in Asia, and "C 039"—a highly
susceptible indica cultivar. They first separated qualitative resistance
from quantitative resistance by identifying RILs showing complete resistance
to each of the five isolates used in the test. The RILs that did not
contain genes governing qualitative resistance were then tested for
quantitative resistance by inoculation under greenhouse conditions with
the isolate P 06-6 in polycyclic tests. Ten chromosomal segments (QTLs)
were found to be associated with effects on lesion numbers. They also
tested the entire RIL populatiou under field conditions at two blast
screening sites in the Philippines and Indonesia and foimd that QTLs
identified in greenhouse tests were good predictors of blast resistance at
the two field sites. A very interesting finding was that three of the
markers reported to be linked to complete resistance in previous studies
were associated with QTLs for partial resistance. This led the authors to
propose that complete resistance and partial resistance may be controlled
by the same genes (alleles), showing different reactions to different
isolates of the fungus. Such a hypothesis implies that a resistance gene

244 Rice Breeding and Genetics: Research Priorities and Challenges
Table 11.1
Genes for disease resistance that have been tagged using molecular markers
■i.:
¡ I I l i i
4' :|i!
li'hlS
Gene Disease Source of Chromo- Closest Distance Reference
resistance some markers (cM)
Xa-1 Bacterial Kogyoku 4 Y5212L 0.9 Yoshimura ei 1996
blight C600 0
Y5212R 1.7
Xa-3 Bacterial Chugoku45 11 XnpblSl 2.3 Yoshimura et al., 1995
blight
Xa-4 Bacterial IR20 11 XNpblSl 1.7 Yoshimura et al., 1995
blight
Xa-5 Bacterial IR1545-339 5 RG556 0.7 Blair and McCouch,
blight RZ390 0.8 1997
Xa-10 Bacterial IRBBIO 11 C^072oo[| 5.3 Yoshimura et al., 1995
blight CD0365 16.2
Xa-13 Bacterial IR66699-5-5-4-2 8 RZ28 4.8 Zhang et al., 1996a
blight RG136 3.7
Xa-21 Bacterial 0. longistaminata 11 AB9 2.8 Williams et al., 1996
blight RG103 0
560 1.8
Xa-22(t) Bacterial Zhachanglong 11 L190 0.8 Lin et al., 1998
blight R1506 0
G3132B 0.7
Pi-1 Blast LAC23 11 RZ536 14.0 Yu et al., 1996
Pi-2(t) Blast 5173 6 RG64 2.8 Yu ei a/., 1991
Pi-4(t) Blast Tetep 12 RG869 15.3 Yu et al., 1991
Pi-5(t) Blast Morobereken 4 RG498

Qifa Zhang and Sibin Yu 245
that exhibits a qualitatively compatible reaction with a fungal isolate
may also provide some protection against the same isolate.
Pertinent data were also obtained by Luo (1998) who studied BLB
resistance of an RIL population from a cross between "Teqing", a
cultivar from China and "Lemont", a cultivar from the United States.
The population and the parents were inoculated with three strains of the
pathogen—CR4j, CR6 , and CX08—and lesion length was used as the
measurement for level of infection. "Teqing" was qualitatively resistant
to CR4 and CX08 but susceptible to CR6 , while 'Xemont" was susceptible
to all three strains. As expected, the Fj was resistant to CR4 and
CX08 and highly susceptible to CR6 .
The lesion length caused by CR4 and CX08 each showed a bimodal
distribution suggesting the involvement of major genes for resistance
although there was also considerable variation in lesion length within
the resistant and susceptible classes. On the other hand, lesion length
caused by CR6 exhibited continuous distribution. Interestingly, transgressive
segregation was observed in both directions, arid a number of
the RILs displayed a highly resistant reaction. Moreover, mapping
analysis revealed a major gene for resistance to all three strains located
on chromosome 11, which was inferred to be Xa4. This gene explained
65.2, 55.2, and 52.1% of the total variation in lesion length caused by
these three strains respectively. In all three cases, the allel from "Teqing"
had a large effect in reducing the lesion length. The study also identified
a number of QTLs for resistance, almost all of them located in close
proximity to loci governing resistance to various diseases. Results from
this study also strongly suggest that QTLs and major genes are different
alleles of the same loci, and even major genes can confer nonspecific
resistance to the pathogen.
Quantitative resistance plays an important role in a system in which
genes for qualitative resistance are not available. This is the case with
the sheath blight disease caused by Rhizoctonia solani. Although shealth
blight is one of the most serious diseases in many rice growing areas,
major gene(s) conferring qualitative resistance has not been found despite
extensive screening in both cultivated gene pools and among wild
relatives. However, considerable variation exists among different varieties
in the level of disease severity, which provides the basis for quantitative
resistance that confers a certain level of protection against R. solani
under field conditions.
Li et at. (1995b) conducted a QTL analysis of quantitative resistance
to K. solani using a population of 255 F4 families from a cross between a
susceptible variety "Lemont" and a resistant (partial resistance) variety
"Teqing". The population was evaluated in two years with artificial

246 Rice Breeding and Genetics: Research Priorities and Challenges
inoculation under field conditions. Analysis identified six QTLs contributing
to R. solani resistance in this population. These six QTLs were
located on six different chromosomes and collectively explain approximately
60% of the genotypic variation in this population. Analysis also
showed that alleles from the resistant parent at five of the QTLs contributed
to reduced level of the disease infection^ but at one of the QTL, the
allele from the susceptible parent produced an increase in resistance.
Such a distribution of alleles conferring a resistance reaction from both
resistant and susceptible parents is very similar to that revealed by QTL
analysis of typical quantitative traits.
i ! E
MAPPING AND GENETIC ANALYSIS OF INSECT RESISTANCE
Rice is the host for a large number of insects. Yield losses due to damage
by insects are severe in all rice-growing areas of the world. In addition to
' the direct damage caused by feeding, some leaf insects, e.g. planthopper
and leaf hopper, are also vectors for viral diseases, leaf transmitting the
pathogens while feeding on the plants. A large number of studies on the
inheritance of insect resistance in rice (Khush and Brar, 1991) have also
identified a large number of varieties with resistance to the various
insects.
A number of studies in mapping genes for insect resistance in rice
have also been undertaken (Table 11.2). The gene conferring GLH
resistance is located at the same position as the one specifying tungro
spherical virus (TSV) resistance (Table 11.1). All the three genes for
Table 11.2 Genes for pest resistance that have been tagged using molecular markers
mu
Gene Pest Source of
resistance
GLH
B ph-l(t)
Bph-W(t)
Bph-?
Gm2
Gm4(t)
Green
leafhopper
Brown
planthopper
Brown
planthopper
Brown
planthopper
Gall
midge
Gall
midge
Chromosome
Closest
markers
Distance
(cM)
Reference
ARC11554 4 RZ262 5.5 Sebastian et a i, 1996
IR28 12 XNpb248 Hirabayashi and
Ogawa, 1995
0 . australiensis 12 RG457 3.68 Ishii et a i , 1994
IR64 12 Sdh-1
RG463
Huang et al., 1997
Phalguna 4 RG329 1.3 Mohan et a i, 1994
ARC6650
RG476 3.4
Abhaya 8 ^^0570 Mohan et a i, 1997

Qifa Zhang and Sibin Yu 247
brown planthopper resistance that were subjected to molecular marker
mapping are mapped to chromosome 1 2 ; the identity for one of them
need further characterization (Huang et ah, 1997). Another observation
is that the precision of the mapping is not quite as good as in the cases of
disease resistance; the linkages of the markers to the genes were not very
tight and markers flanking both sides were not available for most of the
genes mapped. ThuS; more work is needed for identifying tightly linked
markers for marker-assisted selection. More work should also be
conducted for tagging other previously identified genes for insect
resistance.
A major difficulty in tagging insect resistance is the shortage of
genes for resistance to a number of economically very important insects.
For example, the yellow stem borer is one of the most damaging pests in
many rice-growing areas of the world. However, genes for stem borer
resistance are rare in the cultivated gene pools. Future efforts should be
directed toward the identification of the germplasms for resistance to
this Insect; and also to transference of the genes into cultivated
germplasm.
MAPPING AND GENETIC ANALYSES OF GRAIN QUALITY
Grain quality represents one of the major problems of rice production in
many rice-producing areas of the world. There are many components
contributing to rice quality; the most important components are probably
cooking and eating qualities, which involve a number of physical
and chemical characteristics of the starch and are also related to the
appearance of the grains.
The most important constituents of cooking and eating quality are
amylose content, gelling temperature, and gelling consistency. It has
been determined that the waxy gene located on chromosome 6 plays a
major role in controlling amylose content (Wang et ah, 1995; Tan ef ah,
1998). Tightly linked markers to this locus on both sides are available in
the two high density maps (Causse et ah, 1994; Kurata et ah, 1994), It was
also shown recently that gelling temperature and gelling consistency are
likewise controlled by the waxy locus or closely linked chromosomal
block (Table 11.3).
The mo^t important trait for grain appearance is grain length, especially
length-to-width ratio, which to some extent is related to cooking
and eating quality, as this ratio is highly correlated with chalkiness of
the endosperm. Grain length appeared to be inherited as a quantitative
trait in studies in which the grain length of the parents did not differ
much; four to seven QTLs were identified in these studies (Huang et ah,

248 Rice Breeding and Genetics: Research Priorities and Challenges
1997; Redona and Mackill/ 1998). However, a major gene effect was
detected in a cross between two parents that showed a large difference
in grain length (Table 11.3).
Another important characteristic of cooking quality is the cookedgrain
elongation. The study by Ahn et ah (1993) showed that a gene for
this trait is located on chromosome 8 . Ahn et al, (1992) also showed that
a gene for fragance of cooked rice is also located on chromosome 8 with
a loose linkage to the locus for elongation.
Table 11.3
Major genes for grain quality that have been tagged using molecular markers
Gene Trait Source Chromo- Closests Distance Reference
some markers (cM)
Wx
Gl
fg r
Amylase
conter\t
Minghui 63 6 Wx 0 Wang et al., 1995;
Tan et a i, 1998
Gelling Minghui 63 6 Wx Tan et a/,, 1998
temperature
¿elling Minghui 63 6 Wx Tanaf al., 1998
consistency
Grain length
Aroma
Minghui 63 3
Tan et al., 1998
8 RG28 4.5 Ahn et a l, 1992
Grain Basmati 370 8 RZ323 Ahn et al., 1993
elongation
RZ562
MAPPING AND GENETIC ANALYSIS OF FERTILITY RELATED
GENES
Fertility Restoration of WA-CMS
The most widely used cytoplasmic male sterility (CMS) system in hybrid
rice production is the wild-abortive (WA) CMS. It is estimated that
hybrids developed by making use of the WA-CMS accounted for
approximately 90% of the hybrids produced in China in the past. Many
studies demonstrated that two independent loci control fertility
restoration in this system (Zhou, 1983; Young and Virmani, 1984; Li and
Yuan, 1986). However, contradictory results have been obtained in the
literature concerning the chromosomal locations of the two loci. The
trisomic analysis of Bharaj et al (1995) suggested that the two Rf loci
were located on chromosomes 7 and 10. A molecular marker study by
Zhang et al (1997) based oh segregation populations from crosses between
isogenic lines located one of the loci (R/3) on chromosome 1. A
further study by Yao et al (1997), who searched for the Rf loci using
RFLP markers covering the entire rice genome in an F2 population from
a cross between the parents of the most widely used hybrid (Shanyou

Qífa Zhang and Sibin Yu 249
63)^ ''''Zhenshan 97A" and ''Minghui 63"/ showed that the two loci were
located on chromosomes 1 and 10. The locus on chromosome 1 was
located in the same region as R/3 determined by Zhang ef at (1997). But
the identity of the locus on chromosome 10 remained to be characterized,
since there was evidence indicating that another locus for fertility
restoration of the BT CMS system, Rfl, was located in the nearby region
on chromosome 10 (Ichikawa et at, 1997). Yao et at (1997) thus designated
this locus as R/(u) in which "u" indicated uncharacterized identity.
Markers closely linked to and bracketing both Rf3 and R/(u) have
been obtained (Table 11.4).
P h o t o p e r io d -s e n s it iv e g e n ic m a l e s t e r il it y
The photoperiod-sensitive genic male sterility (PSGMS) rice was found
as a spontaneous mutant from a late japónica variety Nongken 58 grown
in Hubei Province, China. Numerous studies conducted in the past had
established that this mutant possesses a number of desirable
characteristics that might be useful in hybrid rice: pollen fertility of this
mutant is regulated by photoperiod length; it is completely sterile when
grown under long-day conditions, whereas pollen fertility varies when
it is grown under short-day conditions. Studies have also demonstrated
that the sterility is controlled by a relatively simple genetic system,
usually one or two Mendelian loci.
However, the results from mapping studies appeared to be complicated.
The study by Zhang et at (1990), who attempted to map the
gene(s) for PSGMS using a series of morphological markers, indicated
that one of the loci segregating for male sterility in their mapping
populations was located on chromosome 5, But in a molecular markerbased
analysis involving an indica PSGMS line, Zhang et at (1994b)
determined that the two loci segregating for male sterility in the population
were located on chromosomes 3 and 7 respectively. They designated
these two loci as pms2 and pmsl according to the amount of effect
that each locus had on fertility. Moreover, in molecular marker-based
analyses of two crosses using Nongken 58S as the PSGMS parent, Mei et
at (1998a) also detected two loci segregating for sterility, one located on
chromosome 7 that was the same as pmsl, and the other on chromosome
12 {pms3). They (Mei et at, 1998b) further determined that pms3 was the
locus on which the original PSGMS mutation occurred that changed the
cultivar Nongken 5S to the PSGMS rice Nongken 58S. Markers bracketing
all three loci were identified although the distance varied from one
case to another (Table 11.4).
Further complications have arisen in the practice of breeding for new
PSGMS lines, especially in indica genetic backgrounds. The two major
difficulties encountered are instability of sterility in many newly bred

250 Rice Breeding and Genetics; Research Priorities and Challenges
indica PSGMS lines, such that variable amounts of seed setting occur
when the temperature fluctuates at certain stages of rice growth and development,
and low reversibility of other newly developed PSGMS lines.
The sterility of these lines is highly stable but the lines cannot revert to
normal fertility under short-day conditions. A molecular marker-based
genetic analysis revealed very complex genetic bases of the stability of
sterility and reversibility to fertility, involving both multiple QTLs and
epistatic interactions between loci for the two phenomena (He etal, 1999).
Thus, molecular markers may be particularly helpful for sorting out the
desired genotypes in the breeding of PSGMS lines.
ii
Table 11.4
Fertility related genes that have been tagged using molecular markers
m II
li It:
Gene Trait* Source Chromosome
Closest
markers
Distance
(cM)
Reference
pmsl PSGMS 32001S 7 RG477 0.5 Zhang et aU, 1994;
R1807 3.8 Wang, 1998
pms2 PSGMS 32001S 3 RG348 1 0 .6 Zhang et aL, 1994
RG191 7.0
ptns3 PSGMS Nongken 58S 1 2 R2708 9.0 Mei et al, 1998b
RZ261 5.5
tmsl TGMS 5460S 8 RZ562 Wang et al., 1995
RG978
tms3(t) TGMS IR32364 6 OP AC6640 7.7 Subudhi et al., 1997
TGMS OPAA7ij5o 1 0 .0
Rfl FR(BT) MTCIOR 1 0 G2155 Ichikawa et al., 1997
Rf3 FR(WA) IR24 1 RG532 2 .6 Zhang effl/., 1997
OPKOSgQQ 5.5 Yao et al, 1997
Rf(u) FR(WA) Minghui 63 1 0 G4004 3.3 Yao et al., 1997
C234 15.2
S5 WC 02428 6 R2429 i.b Liu et al, 1997
RZ450 13.4
PSGMS, Photoperiod sensitive male sterility; TGMS, thermosensitive genic male sterility;
FR, fertility restoration; WC, wide compatibility.
T h e r m o s e n s it iv e g e n ic m a l e s t e r il it y
Three TGMS mutants have been reported in the literature—5460S, H89-
1 and IR32364TGMS—identified in China (Sun et ah, 1989), Japan
(Maruyama et at., 1991), and IRRI (Virmani and Voc, 1991) respectively.
All three TGMS mutants were obtained through irradiation
mutagenesis.
A common characteristic of the TGMS mutants is that their pollen
fertility is regulated by temperature. They are male sterile under high
temperature conditions but show partial to full fertility under low
temperature conditions, although the temperature regime for fertility
induction may not necessarily be the same for these TGMS mutants.
Studies have also shown that, unlike the PSGMS rice, the

Qifa Zhang and Sibin Yu 251
thermosensitive male sterility of all three mutants is controlled in each
by a single recessive gene (Mamyama et al, 1991; Yang et al, 1992;
Borkakati and Virmani^ 1996).
Molecular marker analyses were performed to determine the map
locations of the TGMS genes in two of the mutants (Wang et al, 1995;
Subudhi et al, 1997). In both cases^ the genes were tagged by RAPD
analysis and indirectly mapped by converting the RAPD to RFLP
markers.
W id e c o m p a t ib il it y
Wide compatibility .varieties (WCVs) are a special class of rice
germplasm that is able to producé fertile hybrids whein crossed to both
indica and japónica rice (Ikehashi and Arakh. 1984), while hybrids
between indica and japónica varieties usually show partial sterility
(Kato et al, 1928). The discovery of WCVs brought breeders hope for
breaking the fertility barrier between indica and japónica subspecies and
provided the possibility of exploiting the very strong heterosis
demonstrated in intersubspecific crosses. Results from many studies
(e.g. Ikehashi and Araki, 1986; Liu et al, 1992; Zheng et al, 1992;
Yanagihara et al, 1995) indicated the existence of a gene for wide
compatibility (WCG) at the S5 locus on chromosome 6 that was subsequently
fine mapped by Liu et al (1997) using '02428' as the WCV parent
(Table 11.4). Liu et al (1997) also detected the existence of two minor loci
located on chromosomes 2 and; 1 2 whose joint effect could result in
partial sterility even in the presence of the WCG.
It has also been demonstrated that different WCVs differ in the
genetic basis of wide compatibility. In a study using 'Dular', a variety
from India with a very high level of wide compatibility, as the WCV
parent, Wang et al (1998) identified five loci controlling fertility segregation
in the population, located on chromosomes 1, 3, 5, 6 , and 8 respectively.
The locus on chromosome 6 corresponded to the S5 locus identified
previously. But the locus showing the largest effect was the one on
chromosome 5. They also identified two interactions involving two and
three loci respectively, among the five loci in conditioning hybrid sterility,
indicating a complex genetic basis of wide compatibility.
MOLECULAR MARKER-BASED ANALYSES OF HETEROSIS
Relationship between Molecular Marker Heterozygosity
and Heterosis
One of the major efforts in molecular marker-based study of heterosis
has been concentrated in the characterization of correlation between

252 Rice Breeding and Genetics: Research Priorities and Challenges
Hi
molecular marker heterozygosity and hybrid performance with the hope
of finding a means for predicting hybrid performance using molecular
makers. A number of studies covering a wide range of the cultivated
rice germplasm have been conducted; almost all employed a diallel
design in which the experimental lines were crossed in all possible nonreciprocal
pairs. The parents were assayed using a large number of
molecular markers covering the entire rice genome and genotypes of the
hybrids were deduced from the parental genotypes. All the hybrids and
parents were examined for agronomic performance in replicated Held
trails. Two statistics were adopted to provide measurements for
heterozygosity of the F| genotypes^ general heterozygosity^ and specific
heterozygosity. General heterozygosity measured the level of
heterozygosity based on all markers included in a study and specific
heterozygosity of a trait was based on markers that appeared to have
significant effects on that trait detected using one-way variance arialysis.
Variable results were produced from these studies (Table 11.5). For
example, in a study involving eight elite parental lines widely used in
hybrid rice production in China, Zhang et al (1994a, 1995) detected high
correlations between specific heterozygosity and heterosis for yield and
yield component traits. However, using essentially the same set of
molecular markers and the same analysis, much lower correlations were
observed in the studies of a set of indica varieties and a set of japónica
varieties with a broad range of representations, including parents of
hybrid rice, modern elite cultivars, primitive cultivars and land races
from several countries (Zhang et al, 1996b). On the other hand, very few
correlations were detected in the study involving intersubspecific
crosses making use of the wide compatibility gene (Zhao et al., 1998).
Thus, clearly, the correlations between genotype heterozygosity and
hybrid performance were highly variable depending on the genetic
material used in the studies.
Genetic Basis of Heterosis
There have been two major studies in rice investigating the genetic basis
of heterosis. The study of Xiao et al (1995) involved the use of an F^
population of 194 recombinant inbred lines derived from a cross
between an indica line "9024" and a japónica variety "Lunhui 422". The
RILs were backcrossed to both parents, resulting in a total of 388 BCiFy
lines. All the BC lines, the RILs, the parents and the were examined
for performance of 12 quantitative traits in a replicated field trial, A total
of 37 QTLs were detected at LOD threshold 2.0 in the BC1F7 populations.
Twenty-seven of the QTLs were detected in only one of the two BC
populations; the heterozygotes were superior to the respective

Qifa Zhang and Sibin Yu 253
homozygotes in 82% of the cases. The remaining ten QTLs were detected
in both BC populations and the heterozygotes had phenotypes falling
between the two homozygotes. The authors concluded that dominance
is the major genetic basis of heterosis in this population.
Table 11.5
Correlations between heterozygosity and hybrid performance
in various genetic materials analyzed
Tillers/plant Grains/panicle Grain weight Yield
Parents of elite hybrids (Zhang et ni., 1995)
Performance 0.13/0.10 .0.13/0.18 , 0.53**/0.70** 0.36/0.48**
Heterosis 0.49’^V0.35 0.54**/0.71 0.30/0.25 0.56**/0.77**
Indica mixture (Zhang et al., 1996b)
Performance -0.14/0.27 0.47**/0.57** 0.37*/0.61** 0.43**/0.44**
Heterosis 0.17/0.30 0.24/0.32 0.14/0.54** 0.34**/0.45**
Japónica mixture (Zhang et aL, 1996b)
Performance 0.12/0.52** 0.42*»/0.68** -0.08/0.26 0.48**/0.59**
Heterosis 0.17/-0.03 0.12/0.13 -0.04/-0.05 0.17/0.09
Inlersubspecific crosses (Zhaoef al., 1998)
Indica x indica
Performance -0.55/-0.36 0.44/0.84** 0.25/0.10 0.20/0.26
Heterosis 0.11/-0.16 0.22/Û.33 0.49/0.32 0.37/0.33
Japónica X japónica
Performance 0.28/0.51* -0.15/-0.56* 0.18/0.54* 0.43/0.65**
Heterosis 0.10/0.15 0.02/-0.23 0.03/24 0.41/52*
Indica x japónica
Performance -0.05/-0.21 0.19/0.29 -0.08/-0.12 0.20/0.32
Heterosis 0.27/0.00 0.10/0.27 0.18/-0.08 0.24/-0.11
** significant at 0.05 and 0.01 probability levels respectively. Calculation on general
he lerozygosity/ specific he terozygosity.
I I!
In another study^ Yu et al (1997) investigated the genetic basis of
heterosis of a cross between the parents of an elite rice hybrid, shanyou
63, the most widely grown hybrid in rice production in China. Data for
yield and three traits that were components of yield were collected over
two years from replicated field trials of 250 F2 :3 families from this cross.
Substantial heterosis still existed in the F 3 families for yield and grains
per panicle. A total of 32 QTLs were detected at LOD 3.0 for the four
traits using 150 segregating marker loci covering the entire rice genome.
Twelve of the QTLs were observed in both years and the remaining 20
were detected in only one year. Overdominance was observed for most
of the QTLs for yield and also for a few QTLs for the component traits.
The most striking feature is the observation of frequent and widespread
occurrence of digenic interactions in this population, including additive
by additive, additive by dominance and dominance by dominance types
of interactions. The interactions involved large numbers of marker loci.

254 Rice Breeding and Genetics: Research Priorities and Challenges
most of which were not detectable on a single locus basis. The authors
concluded that epistasis plays an important role both in the inheritance
of quantitative traits and as the genetic basis of heterosis.
There may be many reasons for the disagreement between the two
studies in the conclusions regarding the genetic basis of heterosis^ such
as the types of genetic materials used in the studies and the levels of
heterosis of the crosses. However, the results from both studies clearly
demonstrate that molecular markers, hence genome mapping, have
provided efficient tools for dissecting the genetic basis of heterosis.
ilHi
MAPPING AND GENETIC ANALYSES OF GENES FOR
AGRONOMIC TRAITS
Numerous studies have been conducted on mapping genes of agronomic
traits. Traits studied include those extensively investigated in
breeding programs such as yield and yield component traits, plant
height and heading date, and those not frequently examined in conventional
genetic analyses, such as root morphology, seedling vigor and
stress tolerance which are related to agronomic performance. These
studies demonstrated that most of the traits are quantitatively irrherited
and the genes are thus mostly QTLs.
Yield and Yield Component Traits
Studies are increasing on mapping QTLs for yield and three other traits,
including tillers per plant, grains per panicle and grain weight, that are
components of yield. Figure 11.1 summarizes the results from seven
studies involving eight populations conducted in recent years. A total of
69 QTLs were detected including 16 for yield, 19 for grains per panicle, 9
for tillers per plant, and 25 for grain weight. Ten of the QTLs were
shared in two or more populations, including one for yield, one for
grains per panicle, and eight for grain weight. The remaining QTLs were
found in only one of the populations. It is also clear from Figure 11.1 that
some of the chromosomal regions are much more active than others in
controlling the expression of QTLs; such QTL-active regions appeared
to be concentrated on chromosomes 1, 2, 4, 5, and 8 . In contrast, some
chromosomes, especially chromosomes 9 and 12, did not seem to have
very much effect on yield and yield component traits. Such concentrated
distribution of the QTLs for yield and yield component traits may have
important implications in rice-breeding programs.

7 7 7
Qifa Zhang and Sibin Yu 255
2
RG324
CDO507
Fig. 11.1
iContd)

256 Rice Breeding and Genetics: Research Priorities and Challenge

Qifa Zhang and Sibin Yu 257
6
1ÍZ39Ü
C263
R830
CDO580
BGL760
RG360
CÍ19
R223?
RG671
C1003X
R565
C952
C1084X
R1952
C226a
RZ398
RZ450
RZ588
CD0226
RZ945
R1436
C1239X
CDO105*
RG13
R1553
CD0345
G1314a
R594
C43
CD089
C1402
RG470X
RZ70
C246
RG697X
R1714
RG435
RG1I9
CD054
Fig. 11.1
(Cotttrf)
RG213
R2147, G200
R2I71
CD0544
RG653
RZ405
G342

258 Rice Breeding and Genetics: Research Priorities and Challenges
7 8
C1017
RZ143
C390
R1010
R2285
C1121
R1813
RG1034
RG978

— n
Qifa Zhang and Sibin Yu 259
10
-R C701
•R1933
C148
R1629X
I RZ920
RZ561
C1286
2b
1 1 .
R2825
CD098
R1877
C1361
CDO250
CD094*
C488
RG 134
C16
C809
■RZ421
•C223
■C405*
I centromeric region 1 yield grains/panicle
M tillers/plant
Il 1000-seed weight
Fig. 11.1
(Contd)

260 Rice Breeding and Genetics: Research Priorities and Challenges
n 12
7 7
R Z 5 2 5
" C 3 6 2 X
:i04X
Cl 116a
■R2918X
■C794
•RG118
■G320
■ R G 1 0 9 4
I ------RGI67
'^G257
‘RG16
^RG2
' CD0534
■C1172
■ C 5 0 b
G1CI3
■ G 1 4 6 5
■ C 9 5 0
■ G 1 8 1
" R Z 5 3 6
•C104X
•RG574X
RZ103X
•R2918X
RZ257
' R2672X
R367
' ABC 45 4
R3375
■R1869
RG341
' R887X
■RG9
' RG457
C751a
■RG413*
“ ■ ^ R' R2708 2
^ 'reD 0344*
\ y R G 5 4 3
\ V r G 4 6 3 , R G 9 0 1
X I 069
" W g958
J ^ 9 9 6
\ R 1 6 8 4
\ g181
Fig. 11.1 pisWbuiionofquanliealivetraiilocl(QTU)lbryieldand(heirthreecomponent
ttmts idmti fied in eight popuiations of rice. The most iikeiy positions of the QTLs
( UD > 2.4) for the four traits are placed on the map of Xiong et a i (1998a) that
integrated markers from the maps of Causse et ul. (1994) and Kurata et al., (1994).
of each bar represents the population used for QTL detection;
l-z55^ ^2:4 lines from the Lemont/Teqing cross (Li et al., 1997); 2a--171 F,
plants from Waiyin 2/CB (Lin et al,
AoiT c ^ r ^ ooobied haploid lines from Zhaiyeqing/Jingxi 17 (Lu et al, 1996);
4-231 F2 plants from Palawan/IR 42 (Wu et al, 1996); 5-194 recombinant inbred
lines from 9024/LH422 (Xiaoef al, 1996a); 6-300 BC2 testcross families of V20/O.
ra/ipo^oii/Ce64 (Xiao ei al, 1996b); 7-250 F2.3 lines from Zhenshan 97/Minghui
63 (Yu c?f al, 1997).

Qifa Zhang and Sibin Yu 261
Plant Height
Two genes for semidwarfism, sdl and sdg^ were mapped to chromosomes
1 and 5 respectively (Cho et at, 1994; Liang ei al, 1994). Eleven
dwarfing genes were identified and mapped to various chromosomal
locations (Table 11.6). A number of QTLs were also detected for this trait
which involved all 12 rice chromosomes (Lu et al, 1996; Xiao et al,
1996a; Huang et al., 1996; Zhuang et al, 1997).
To gain insights into the nature of the major genes and QTLs controlling
plant height, Huang et al (1996) studied the genetic effects of the
loci contributing to plant height usirig five mapping populations. The
difference in height between the two parents was large for all the
populations, presumably due to the existence of major genes for either
dwarfism or semidwarfism. Analysis detected a total of 23 QTLs with
eight of them shared by at least two of the five populations. A very
interesting feature revealed by this analysis is that, for each of the
previously identified dwarfing or semidwarfing genes, at least one QTL
mapped to its close proximity. The authors interpreted such positional
correspondence between the QTLs and major genes for plant height as
evidence for supporting the hypothesis that QTLs and major genes are
different alleles of the same loci.
Heading Date
Heading date in rice is mainly determined by two factors—duration of
basic vegetative growth and photoperiod sensitivity—;cach of which is
Table 11.6
Major gene loci for plant height and heading date that were directly or
indirectly tagged using molecular markers
I
Gene Chromosome Linked marker Reference
sdl 1 RZ730-RG690 Adapted from Huang et al., 1997
dlO 1 RG345~RZ276
dl8 1 RG472
dS 2 RG654-RG256-RG95
d32 2 RZ318-RG157
d30 2 RG157-RG171
d56 3 RG104-RG144
d31 4 RZ590-RG214
d ll 4 RG163 (or RG218)
sdg 5 RG403-RG13-CD0105
d9 6 RG648-RG424
d27 11 RG103
d33 12 RG463-RG457
Hdl 6 R1679 Yano et«/., 1997
Hd8 8 RG333-C1121 Lu et al., 1996; Xiong et al,, 1998b

262 Rice Breeding and Genetics: Research Priorities and Challenges
controlled by several genes (Kinoshita^ 1995). One of thé major genes for
photoperiod sensitivity, Sel, located on chromosome 6 was tagged'by
Mackill et al. (1993), which was further resolved by Yano et al. (1997)
using QTL analysis as explaining 67% of the total phenotypic variation.
This locus was located on top of the marker locus R1679 in the map of
Kurata et al. (1994), Scrutiny of the published QTL analyses (Li et al,
1995a; Lu et aï-, 1996; Xiao et ai, 1996a; Xiong ef al, 1998) also indicated
the existence of a locus on chromosome 8 that had a major effect on
heading date by explaining from 33% to 52% of the phenotypic variation.
This is likely to be another locus for photoperiod sensitivity. In addition,
a number of QTLs with minor effects pn heading date were also detected
in various studies were which involved chromosomes 1, 3, 6 , 7, 8 , 10,
and 1 1 .
Root Morphology and Stress Tolerance
Several studies have been conducted in QTL mapping of root morphology
in the context of drought tolerance (Table 11.7). Very large
numbers of QTLs were detected for some of the traits, which indicated
that the genetic basis for root morphology is complex. A comparison of
the results obtained from two different populations (Ghampoux ei al,
1995 vs. Yadav et al, 1997) showed that 1 to 3 of the QTLs for the various
traits could be repeatedly detected. However, none of the QTLs
exhibited a major effect.
Table 11,7 Summary of the QTLs detected for root morphology traits
Trait Number Method Statistical Variance Reference
of QTLs of threshold explained
detected analysis* (%)
Total root number 19 M/Q LOD4.1 8-15 Ray ei al, 1996
Penetrated roots 4 M/Q LOD3.2 6 -8
Root penetration 6 M/Q LOD3.9 7-13
index
Total root weight 6 ANOVA P

Qifa Zhang and Sibin Yu 263
Some studies mapping genes for tolerance to abiotic stress including
submerging and ferrous iron toxicity (Table 11.8) have been done. A
major gene locus, Subl, was identified for submerging tolerance (Xu and
Mackill, 1996). One of the loci detected for ferrous iron toxicity tolerance
also appeared to have a large effect as it explained 32% of the variation
in leaf bronzing (Wu et at, 1997).
Table 11.8
Stress tolerance genes that have been tagged by molecular markers
Subi
Gene Trait Source Chromosome
Submerging
tolerance
Ferrou
toxicity
tolerance
Closest
markers
FA13-A 9 C1232
RZ698
Azucena 1 RG345-
RG381
Distance
(cM)
Reference
4 XÙ and Mackill,
1996
2.3 Wueffl/,, 1997
PERSPECTIVE
Tlie results of the extensive studies in tagging and genetic analyses of
genes will have significant impacts on rice genetic improvement.
Marker-assisted Selection
Recently, molecular marker-assisted selection has been successfully applied
in rice-breeding programs. For example, Huang et al. (1997) used
marker-assisted selection to pyramid multiple genes for resistance to
BLB, which increased both the spectrum and level of resistance of the
resultant lines. Chen et aL (1998) developed a new version of Minghui
63, the best restorer line in Chinese hybrid rice production, by precise
replacement of the Xa21 containing region. The improved version contained
only a fragment of less than 4.0 cM (0.35%) surrounding the Xa21
locus from the donor parent with the remainder (>99.65%) of the genome
from the recipient. Such a precise replacement may be very Important
for maintaining the combining ability of the restorer line.
It is expected that molecular marker-assisted selection will have a
major role to play in future in genetic improvement of crops, including
rice. This is not only because the technique itself has provided a highly
efficient tool for speedy and precise selection, but also because it
possesses several advantages compared to genetic transformation. First,
it does not require isolation of the targeted gene, which often takes years
and considerable resources to accomplish. Second, most of the gene
constructs, such as those commonly used in many transformation

264 Rice Breeding and Genetics: Research Priorities and Challenges
studies, are now covered by intellectual property rights and hence are
not freely available for varietal development. Third, the progeny developed
by marker-assisted selection in general does not suffer from the
adverse effects such as over- or underexpression, transgene silencing
etc., which are now frequently reported with transgenic plants. Thus the
performance of the progeny resulting from marker-assisted selection is
much more predictable than that from transformation. The large number
of genes that have been precisely tagged and mapped will therefore
provide a rich source for marker-assisted breeding.
Gene Isolation
11
I
Currently, the most common practice for obtaining new genes is mapbased
cloning. Molecular markers that are closely linked to genes of
interest can serve as the starting points for cloning the genes following
the map-based cloning approach. Using this approach, several important
genes have now been isolated in rice, including two genes for BLB
resistance (Song et ah, 1995; Yoshimura et ah, 1998) and one gene for
blast resistance (Wange et ah, 1998). It is also known that molecular
cloning of a number of genes is in progress (e.g., Lin et al., 1998; Lu et aL,
1998). It can be expected that the process of gene isolation using this
approach will be greatly accelerated with the advances of the
international effort in sequencing the entire rice genome, which
supposedly will be completed in the next 5 to 10 years. It is highly likely
that all the genes that are accurately mapped with closely linked markers
could be quickly isolated with the availability of the sequence
information.
Recent developments in DNA-chip technologies may also provide a
powerful tool for large-scale isolation of new gens in the near future
(Lemieux et al., 1998). It can be expected that large numbers of genes will
become available for rice improvement in the next decade.
Germplasm Enhancement
Wild relatives have for long been recognized as important gene pools for
qualitatively inherited resistance and/or tolerance to biotic and abiotic
stress. Recent studies using molecular markers have demonstrated that
wild relatives may also be important sources of useful genes for
agronomic performance including yield and yield component traits. For
example, Xiao et al. (1996b) reported two QTLs from a wild rice that
showed large effects in increasing the performance of an elite rice
hybrid. Such a finding has generated considerable interest in identifying

I—^
Qifa Zhang and Sibin Yu 265
genes for agronomic performance from wild relatives that are potentially
useful for varietal improvement. Indeed, in several studies, DNA
segments from wild relatives have been reported to have significant
effects on yield and yield component traits in the genetic backgrounds of
cultivars, which were referred as wild QTLs. Such wild QTLs may be
potentially rich source of genes of agronomic importance.
In summary, recent developments in genome mapping and genetic
engineering have provided a knowledge base, identified germplasm
resources, provided useful genes, and offered effective tools for rice
improvement. Integration of this knowledge and the genetic resources
into breeding programs will greatly increase the efficiency of varietal
development.
References
Ahn, S.N., Bollich, C.N. and Tanksley, S.D. 1992. RFLP tagging of a gene for aroma in rice
Theor. Appl 84: 825-828.
Ahn, S.N., Bollich, C.N., McClung, A.M. and Tanksley, S.D. 1993. RFLP analysis of genomic
regions associated with cooked-kernel elongation in rice.Theor. Appl. Gen. 87:27-32.
Bharaj, T.S., Virmani, S. S. and Khush, G.S, 1995, Chromosomal location of fertility restoring
genes or 'wild abortive' cytoplasmic male sterility using primary trisomiqs in rice.
Euphytica 83:169-173.
Blair, M.W. and McCouch, S.R. 1997. Microsatellite and sequence-tagged site markers diagnostic
for the rice bacterial leaf blight resistance genexa-5. Theor. Appl. Gen. 95:174-184.
Borkakati, R.P. and Virmani, S. S. 1996. Genetics of thermosensitive male sterility in rice.
Euphytica 8 8 :1-7.
Causse, M.A., Fulton, T.M., Gho, Y.G., Ahn, S. N., Chunwongse, J., Wu, K.S., Xiao, J.H., Yu,
Z.H., Ronald, P.C., Harrington, S.E., Second, G., McCouch, S.R. and Tanksley, S.D. 1994.
Saturated molecular ihap of the rice genome based on an interspecific backcross population.
Genetics 138:1251-1274.
Champoux, M.C., Wang, G,, Sarkarung, S., Mackill, D.J., O'Toole, J.C,, Huang, N. and
McCouch, S.R, 1995. Locating genes associated with root morphology and drought
avoidance in rice via linkage to molecular markers. Theor. Appl. Gen. 90:969-981.
Chen, S., Lin, X., Xu, C. and Zhang, Q. 1998. Marker—assisted selection to improve bacteria
blight resistance of Minghui 63, an elite restorer of hybrid rice. Abstracts. 18th Inti. Gen.
Cong. 10-15 August 1998, Beijing, p. 158.
Cho, Y.G., Bun, M.Y., McCouch, S.R. and Chae, Y.A. 1994. The semidwarf gene, sd~l, of rice
{Oryza sativa L,), II, Molecular mapping and marker-assisted selection, Theor. Appl. Gen.
89: 54-59.
Hayano-Saito, Y., Tsuji, T., Fujii, K., Sai to, K., Iwasaki, M. and Sai to, A. 1998. Localization of
the rice stripe disease resistance gene, $tv-h', by graphical genotyping and linkage analysis
with molecular markers, Theor, Appl. Gen. 96:1044-1049.
He, Y., Yang, J., Xu, C., Zhang, Z. and Zhang, Q. 1999. Genetic bases of instability of male
sterility and fertility reversibility in photoperiod sensitive genic male sterile rice. Theor,
Appl. Gen. in press.
Hirabayashi, H. and Ogawa, T. 1995. RFLP mapping of Bph-1 (brown planthopper resistance
gene) in rice. Breed. Sci. 45:367-^91.

'i I
266 Rice Breeding and Genetics: Research Priorities and Challenges
Huang, N., Courtois, B,, Khush, G.S., Lin, H.X., Wang, G,, Wu, P. and Zheng, K. 1996'.
Association of quantitative trait loci for plant height with major dwarfing genes in rice.
Heredity 77: 130-137,
Hüang, N., Angeles, E.R., Domingo,]., Magpantay, G., Singh, S,, Zhang, G., Kumaravadivel,
N., Bennett, J, and Khush, G.S. 1997. Pyramiding of bacterial blight resistance genes in
rice: marker-assisted selection using RFLP and PCR. Theor. AppL Gen. 95:313-320,
Huang, N,, Parco, A., Mew, T., Magpantay, G., McCouch, S.R., Guiderdoni, E., Xu, J.,
: Subudhi, P., Angeles, E,R. and Khush, G.S, 1997, RFLP mapping of isozymes, RAPD and
QTLs for grain shape, brown planthopper resistance in a doubled haploid rice population,
Moke. Breed 3; 105-113,
Ichikawa, N., Kishimoto, N., Inagaki, A., Nakamura, A,, Koshino, Y., Yokozeki, Y., Oka, M.,
S'amoto, S., Akagi, H., Higo, K., Shiniyo, C., Fujimura, T. and Shimada, H. 1997. A rapid
PCR-aided selection of a rice line containing the Rf-1 gene which is involved in restoration
of the cytoplasmic male sterility. Moke, Breed. 3:195-202,
Ikehashi, H. and Araki, H. 1984. Variety screening of compatibility types revealed in Fj
fertility of distant crosses In rice. Jap. /, Breed. 34:304-^13,
Ikehashi, H. and Araki, H. 1986. Genetics of F^ sterility in remote crosses of rice. In: Rice
Genetics. Symp, 27-31 May 1985. IRRI Manila, Philippines, pp, 119-130.
Inukai, T., Nélson, R.J., Zeigler, R.S., Sarkarung, S., Mackill, D.J., Bonman, J.M., Takamure, I.
and Kinoshita, T. 1994. Allelism of blast resistance genes in near-isogenic lines of rice.
Phytopath. 84: 1278-1283.
Ishii, T., Brar, D.S., Multani, D.S. and Khush, G.S. 1994. Molecular tagging of genes for brown
planthopper resistance and earliness introgressed from Oryza australiensis into cultivated
rice, O, sativa. Genome 37: 217-221.
Kato, S., Kosaka, H. and Kara, S. 1928. On the affinity of rice varieties as shown by fertility of
hybrid plants. Bull Set, fac. Agríe, Kyushu Univ. 3; 132-147.
Khush, G.S. and Brar, D.S. 1991. Genetics of resistance to insects in crop plants. Adv. Agron.
‘ 45: 223-264.
Kinoshita, T. 1995. Report of commitee on gene symbolization, nomenclature and linkage
groups. Rice Genet. Newslett. 12; 9-153.
Kurata^ N., Nagamura, Y., Yamamoto, K., Harushima, Y., Sue, N., Wu, J., Antonio, B.A.,
ShomUra, A., Shimizu, T., Lin, S.Y., Inoue, T,, Fukuda, A., Shimano, T., Kuboki, Y.,
Toyama, T., Miyamoto, Y., Kirihara, T., Hayasaka, K., Miyao, A., Monna, L,, Zhong, H.S.,
Tamura, Y., Wang, Z.X., Momma, T., Umehara, Y., Yana, M., Sasaki, T. and Minobe, Y.
1994. A 300 kilobase interval genetic map of rice including 883 expressed sequences.
Nature Gen. 8 ; 365-372.
Lemieux, B,, Aharoni, A. and Sehena, M. 1998. Overview of DNA chip technology. Molec,
. Breed. 4: 277-289.
Li, Y. and Yuan, L. 1986. Genetic analysis of fertility restoration in male sterile lines of rice.
In: Rice Genetics. IRRI, Manila, Philippines, pp. 617-632.
Li, Z.K., Pinson, S.R.M., Stansel, J.W. and Park, W.D. 1995a. Identification of two major genes
and quantitative trait loci (QTLs) for heading date and plant height in cultivated rice
’ (Oryza sativa L.). Theor. AppL Gen. 91:374-381.
Li, Z.K., Pinson, S.R.M., Marchetti, M.A., Stansel, J.W. and Park, W.D. 1995b. Characterization
of quantitative trait loci (QTLs) in cultivated rice contributing to field resistance to
. sheath blight (Rhizoctonia solani). Theor. AppL Gen. 91:382-388.
Li, Z.K., Pinson, S.R.M., Park, W.D., Paterson, A.H. and Stansel, J.W. 1997. Epistasisfor three
grain yield components in rice (Oryza sativa L.). Genetics 145:453-465.
Liang, C., Gu, M., Pan, X., Liang, G. and Zhu, L, 1994. RFLP tagging of a new semidwarfing
gene in rice. Theor. AppL Gen. 88:898-900.

Qifa Zhang and Sib’in Yu 267
■'4
Lin, H., Qian, H., Zhuang, J., Lu, J., Min, S., Xiong, Z., Huang, N. and Zheng, K. 1996 a. RFLP
mapping of QTLs for yield and related characters in rice {Oryza sativa L.). Theor, Appl.
Gen. 92: 920-927.
Lin, X., Zhang, D., Xie, Y,, Gao, H. and Zhang, Q. 1996b. Identification and mapping a new
gene for bacterial blight resistance in rice based on RFLP markers. Phytopath. 86:1156-
,1159.
Lin, X., Wang, C., Xu, X., Wen, G., Zhang, D., Xie, Y., Peng, K. and Zhang, Q. 1998. Progresses
in map-based cloning of Xfl22{t), a new gene for bacterial blight resistance in rice.
Abstracts 18th Gen. Cong. 10-15 August 1998, Beijing, p. 196.
Liu, A., Zhang, Q. and Li, H. 1992. Location of a gene for wide-compatibility in the RFLP
linkage map. Rice Gent. Newslett. 9; 134-136.
Liu, K., Wang, J., Li, H., Xu, C., Liu, A., Li, X. and Zhang, Q, 1997. A genome-wide analysis of
wide compatibility in rice and the precise location of the locus in the molecular map.
Theor. Appl. Gen. 95:809-814.
Lu, C., Shen, L., Tan, Z., Xu, Y., He, P., Chen, Y. and Zhu, L. 1996. Comparative mapping of
QTLs for agronomic traits of rice across environments using a doubled haploid
population. Theor. Appl. Gene. 93:1211-1217.
Lu, C., Zhai, W., Gong, F,, Li, X., Tian, W-, Zhou, Y., Yang, M., Zhang, Q., Bennetzen, J. and
Zhu, 1*. 1998. Molecular cloning of a R gene homologous sequence linked to the rice
bacterial blight disease resistance gene, Xh7. Abstracts 18th Gen. Cong. 10-15 August
1998, Beijing, p. 39.
Luo, L. 1998. Developing high yield hybrids, probing the genetic basis of heterosis and
bacterial blight resistance in rice. Ph. D. Diss., Huazhong Agricultural Univ.
Mackill, D.J., Salam, M.A., Wang, Z.Y. and Tanksley, S.D. 1993. A major photoperiodsensitive
gene tagged with RFLP and isozyme markers in rice. Theor. Appl. Gen. 85:536-
540.
Maruyama, K., Araki, H, and kiato, H, 1991. Thermosensitive genetic male sterility induced
by irradiation. In: Rice Genetics II. Proc. Second Inti. Rice Gene. Sympo. 14-18 May 1990.
IRRI, Manila Philippines.
McCouch, S.R., Nelson, R.J„ Tohme, J. and Zeigler, R.S. 1994. Mapping of blast resistance
genes in rice. In: Rice Blast Disease R.S. Zeigler, S.A. Leong and P.S. Teng (eds). CAB Inti.
Willingford, UK, pp. 167-186.
Mei, M., Dai, X,, Xu, C. and Zhang, Q. 1998a. Mapping and genetic analysis of the genes for
photoperiod-sensitive genic male sterility in rice using the original mutant Nongken 58S.
Crop Sci. (in press).
Mei, M., Chen, L., Zhang, Z., Li, Z., Xu, C. and Zhang, Q. 1998b. pms3 is the locus causing the
original photoperiod-sensitive male sterility mutation of 'Nongken 58S'. Science in China
(Scries C) in press.
Miyamoto, M., Ando, I., Rybka, K., Kodama, O. and Kawasaki, S. 1996. High resolution
mapping of the indica-derived rice blast resistance genes, I. Pi~b. Molec. Plant-Microbe
Interact. 9: 6-13.
Mohan, M,, Nair, S„ Bentur, J.S., Rao, U.P. and Bennett, T. 1994. RFLP and RAPD mapping of
rice Gm2 gene that confers resistnace to biotype 1 of gall midge (Orseolia oryzae). Theor.
App. Gen. 87: 782-788.
Mohan, M., Sathyanarayan, P.V., Kumar, A., Srivastava, M.N. and Nair, S. 1997. Molecular
mapping of a resistance-specific PCR-based marker linked to a gall midge resistance
gene (Gmit) in rice. Theor. Appl. Gen. 95.777-782.
Naqvi, N.I. and Chattoo, B.B. 1996. Development of a sequence characterized amplified
region (SCAR) based indirect selection method for a dominant blast-resistance gene in
rice. Genome 39:26-30.

268 Rice Breeding and Genetics: Research Priorities and Challenges
Ray, J.D,, Yu, L,, McCouch, S.R., Champoux, M,C., Wang, G. and Nguyen, H.T. 1996. Mapping
quantitative trait loci associated with root penetration ability in rice {Oryza saliva
L.). Theor. Appl. Gen. 92; 627-636.
Redona, E.D, and Mackill, D J. 1998. Quantitative trait locus analysis for rice panicle and
grain characteristics. Theor. Appl. Gene. 96: 957-963.
Rybka, K., Miyamoto, M., Ando, I,, Sai to, A. and Kawasaki, S. 1997. High resolution mapping
of the indica-derived rice blast resistance genes, Il.Pi-fa^.and Pi-ta and a consideration of
their origin. Molec. Plant-Microbe Interac. 10:517-524.
Sebastian, L.S., Ikeda, R., Huang, N., Imbe, T., Coffman, W.R. and McCouch, S.R. 1996.
Molecular mapping of resistance to rice tungro spherical virus and green leafhopper.
Phyiopath. 8 6 ; 25-30.
Song, W.Y., Wang, G.L., Chen, L.L., Kim, H.S., Pi, L.Y., Holsten, T., Gardner, J., Wang, B,,
Zhai, W.X., Zhu, L.H., Fauquet, C. and Ronald, P. 1995. A receptor kinase-like protein
encoded by the rice disease resistance gene, Xa21. Science 270:1804-1806.
Subudhi, P.K., Borkakati, R.P., Virmani, S.S. and Huang, N. 1997. Molecular mapping of a
thermosensitive genetic male sterility gene in rice using bulked segregant analysis.
Genome 40:188-194.
Sun, Z.X., Min, S.K. and Xiong, Z.M. 1989. A temperature-sensitive male sterile line found in
rice. Rice gen. Newslett. 6:116-117.
Tan, Y., Yu, S., Li, J., Xu, C. and Zhang, Q. 1998. The three traits of cooking and eating quality
of rice are specified by the waxy locus. Abstract 6 th Inti. Symp. Rice Molecular Biology,
October 31-November 2, Shanghai and Hangzhou, 1998, p. 63.
Virmani, S.S. and Voc, P.P. 1991. Induction of photo and thermo-sensitive genic male
sterility in indica rice. Agron. Abstr. 119.
Wang, B., Xu, W., Wang, J. Wu, W., Zheng, H., Yang, Z., Ray, J.D. and Nguyen, PI.T. 1995.
Tagging and mapping the thermo-sensitive genic male-sterile gene in rice (Oryza saliva
L.) with molecular nr\arkers. Theor. Appl. Gene. 91:1111-1114.
Wang, F. 1998. PSMS gene pmsl: fine mapping, preliminary physical map construction and
genetic analysis. Ph. D. Diss. Huazhong Agricultural Univ.
Wang, G., Mackill, D.J., Bonman, J.M., McCouch, S.R., Champoux, M.C. and Nelson, R.J.
1994. RFLP mapping of genes conferring complete and partial resistance to blast in a
durably resistant rice cultivar. Genetics 136:1421-1434,
Wang, J., Liu, K., Xu, C., Li, X. and Zhang, Q. 1998. The high level of wide-compatibility of
variety 'Dular' has a complex genetic basis. Theor. Appl. Gene. 97:407-412.
Wang, Z., Zheng, F., Shen, G., Gao, J., Snusted, D.P., Li, M., Zhang, J. and Hong, M. 1995.The
amylose content in rice endosperm is related to the post-translational regulation of the
waxy gene. Plant J. 7: 613-622.
Wang, Z., Yanu, M., Yamanouchl, U., Iwamoto, M., Monna, L., Hayasaka, H., Katayose, Y.
and Sasaki, T. 1998. Map-based cloning and characterization of the rice blast resistance
gene Pib. Abstract 6 th Inti. Symp, Rice Molecular Biology, October 31 -November 2,
Shanghai and Hangzhou, 69. pp.
Williams, C.E., Wang, B., Holsten, T.E., Scambray, J., De Assis, G.D.S.F. and Ronald, P.C.
1996. Markers for selection of the riceX«21 disease resistance gene. Theor. Appl, Genet. 93:
1119-1122.
Wu, P., Zhang, G. and Huang, N. (1996). Identification of QTLs controlling quantative
characters, in rice using RFLP markers. Euphytica 89: 349-354,
Wu, P., Luo, A., Zhu, L., Yang, J., Huang, N. and Senadhira, D. 1997. Molecular markers
linked to genes underlying seedling tolerance for ferrous iron toxicity. In; Plant Nutrition
for Sustainable Food Production and Environment. T. Ando et al. (eds.), Kluwer Academic
Publishers, Japan.
If ii

Qifa Zhang and Sibih Yu 269
Xiao, Li, J., Yuan, L. and Tanksley, S.D. 1995. Dominance is the major genetic basis of
heterosis in rice as revealed by QTL analysis using molecular markers. Genetics 140:745-
754.
Xiao, J., Li, J., Yuan, L. and Tanksley, S.D, 1996a. Identification of QTLs affecting traits of
agronomic importance in a recombinant inbred populatin derived from a subspecific rice
cross. Theor. Appl Gen. 92:230-244.
Xiao, J., Grandillo, S., Ahn, S.N., McCouch, S.R., Tanksley, S.D., Li,and Yuan, L. 1996b.
Genes from wild rice improve yield. Nature 384:223-224.
Xiong, L., Wang, S., Liu, K., Dai, X., Saghai Maroof, M.A., Hu, J. and Zhang, Q. 1998a.
Distribution of microsatelUte and AFLP markers on molecular linkage map in rice. Acta
Botánica Sínica 40; 605-614.
Xiong, L., Liu, K., Dai, X., Xu, C. and Zhang, Q, 1998b. Identification of genetic factors
controlling domestication-related traits of rice using a Fj population of a cross between
Oryza sativa and O. rufipogon. Theor. Appl. Gen. (in press),
Xu, K, and Mackill, D.J. 1996. A major locus for submergence tolerance mapped on rice
chromosome 9, Moke. Breed. 2:219-224.
Yadav, R., Courtois, B., Huang, N. and Melaren, G. 1997. Mapping genes controlling root
morphology and root distribution in a doubied-haploid population of rice. Theor. Appl.
Genet. 94; 619-632.
Yanagihara, S., McCouch, S.R., Ishikawa, K., Ogi, Y., Maruyama, K., and Ikeshashi, H. 1995.
Molecular analysis of the inheritance of the S-5 locus conferring wide compatibility in
indica/japonica hybrids of rice (Oryza sativa L.) Theor. Appl. Genet. 90:182-188.
Yang, R., Liang, K., Wang, N. and Chen, S. 1992. A recessive gene in Indica rice 5460s for
thermosensitive genic male sterility. Rice Gen. Newslett. 9:56-57.
Yano, M. and Sasaki, T. 1997. Genetic and molecular dissection of quantitative traits in rice.
Plant Molec. Biol. 35:145-153.
Yano, M., Harushima, Y., Nagamura, Y., Kurata, N., Minobe, Y. and Sasaki, T. 1997. Identification
of quantitative trait loci controlling heading date in rice using a high-density
linkage map. Theor. Appl. Genet. 95:1025-1032.
Yao, F., Xu, C., Yu, S., Li, J., Gao, Y., Li, X. and Zhang, Q. 1997. Mapping and genetic analysis
of two fertility restorer loci in the wild-abortive cytoplasmic male sterility system of rice
{Oryza sativa L.). Euphytica 98:183-187..
Yoshimura, S., Umehara, Y,, Kurata, N,, Nagamura, Y., Sasaki, T., Minobe, Y. and Iwata, N.
1996. Identification of a YAC clone carrying the Xa-1 allele, a bacterial blight resistance
gene in rice. Theor. Appl. Gen. 93:117-122.
Yoshimura, S., Yoshimura, A., Iwata, N., McCouch, S.R., Abenes, M.L., Baraoidan, M.R.,
Mew, T.W. and Neslon, R.J. 1995. Tagging and combining bacterial blight resistance
genes in rice using RAPD and RFLP markers. Molec. Breed. 1:375-387.
Yoshimura, S., Yamanouch, U., Katayose, Y., Toki, S,, Wang, Z.X., Kono, I., Kurata, N., Yano,
M., Iwata, N. and Sasaki. T. 1998. Expression of Xa-1, a bacterial blight-resistance gene in
rice, is induced by bacterial inoculation, Proc. Natl. Acad. Sci. USA 95:1663-1668.
Young, J.B. and Virmani, S.S. 1984. Inheritance of fertility restoration in a rice cross. Rice Gen.
Newslett. 1; 102-103.
Young, N.D, 1996. QTL mapping and quantitative disease resistance in plants. Ann, Rev.
Phytopath. 34: 479-501.
Yu, S., Li, J., Xu, G., Tan, Y., Gao, Y., Li, X., Zhang, Q. and Saghai Maroof, M.A. 1997.
Importance of epistasis as the genetic basis of heterosis in an elite rice hybrid. Proc. Natl,
Acad. Sci. USA 94: 9226-9231.
Yu, Z., Mackill, D.J., Bonman, J.M. and Tanksley, S.D. 1991. Tagging genes for blast resistance
in rice via linkage to RFLP markers. Theor. Appl. Gen. 81:471-476.

270 Rice Breeding and Genetics: Research Priorities and Challenges
Yu, Z., Mackill, D.J., Bonman, J.M., McCouch, S.R., Guiderdoni, E., Notteghem, J.L. and
Tanksley, S.D. 1996. Molecular mapping of genes for resistance to rice blast {Pyricularia
grísea Sacc.) Theor. Appl. Gen. 93: 859-863.
Zhang, D., Deng, X., Yu, G., Lin X., Xie, Y. and Li, Z. 1990, Chromosomal location of the
photoperiod sensitive male sterile gene in Nongken 58S./. Huazhong Agricu. Uni. 9:407-
419.
Zhang, G., Angeles, E.R., Abenes, M.L.P., Khush, G.S, and Huang, N. 1996a. RAPD and RFLP
mapping of the bacterial blight resistance gene xa-13 in rice. Theor. Appl. Gen. 93:65-70.
Zhang, G., Bharaj, T.S., Lu, Y., Virmani, S.S. and Huang, N. 1997. Mapping of the'R/-3 nuclear
fertility-restoring gene for WA cytoplasmic male sterility in rice using RAPD and RFLP
markers, Theor. Appl Gen. 94; 27-33.
Zhang, Q., Gao, Y., Yang, S., Ragab, R.A., Saghai Maroof, M.A, and Li, Z, 1994a. A diallel
analysis of heterosis in elite hybrid rice based on RFLPs and microsatellites. Theor. Appl,
Gen. 89:185-192,
Zhang, Q., Shen, B., Dai, X., Mei, M., Saghai Maroof, M.A, and Li, Z..1994b. Using bulked
extremes and recessive class to map genes for photoperiod-sensitive genic male sterility
in rice. Proc. Natl Acad. Sd. USA 91: 8675-8679.
Zhang, Q., Gao, Y., Saghai Maroof, M.A., Yang, S. and Li, J. 1995. Molecular divergence and
hybrid performance in rice. Molec. Breed. 1:133-142.
Zhang, Q,, Zhou, Z,, Yang, G„ Xu, C., Liu, K. and Saghai Maroof, M.A, 1996b. Molecular
marker heterozygosity and hybrid performance in indica and japónica rice. Theor. Appl.
Gen. 93; 1218-1224.
Zaho, M., Li, X., Yang, J., Xu, C., Hu, R., Liu, D. and Zhang, Q. 1998. Relationship between
molecular marker heterozygosity and hybrid performance in intra-and inter-subspecific
crosses of vice. Plant Breed, {in press):
Zheng, K., Shen, P., Qian, H. and Wang, J. 1992. Tagging genes for wide compatibility in rice
via linkage to RFLP markers. Cliittese /. Ric. Sci. 6:145-150.
Zheng, K., Qian, H., Zhuang, J., Lu, J., Lin, H. and Kochert, G. 1995. Tagging rice blast
resistance genes via DNA markers. Acta Phytopathologica Sínica 25:307-313.
Zhou, T. 1983. Analysis of R genes in hybrid indica rice of 'WA' type. Acta Agronómica Sínica
9: 241-247.
Zhu, L., Xu, J., Chen, Y., Ling, Z., Lu, C. and Xu, Y. 1994. Mapping an unknown gene for rice
blast resistance using molecular markers. Science in China (Series B) 24:1048-1052.
Zhuang, J., Lin, H., Lu, J., Qian,H., Hittalmani, S., Huang, N. and Zheng, K. 1997. Analysis of
QTL X environment interaction for yield components and plant height in rice. Theor.
Appl Gen. 95\ 799-808.
\ I!

12
Exploitation of Alien Species
in Rice Improvement-
Opportunities,
Achievements and Future
Challenges
K.K. Jena^ and G.S. Khush^
INTRODUCTION
Rice (O. sativa L.) is one of the most important cereal crops in the world
and is the major staple for 35% of the world population. It is cultivated
worldwide under diverse agroclimatic conditions. However, rice
productivity is affected by several biotic (diseases and insects) and abiotic
(adverse soils, temperature and water conditions) stresses limiting
increased rice production. Rapid population growth and shrinking
cultivable land demand increased rice yield by incorporating
agronomically important genes such as resistance to major diseases and
insects from alien species. Some of the major diseases and insects affecting
rice production include bacterial leaf blight (BB), blast (Bl), sheath blight
(ShB), brown planthopper (BPH), white-backed planthopper (WBPH),
^Biotechnology Centre, Mahyco Research Foundation Hyderabad 500 073 India.
^ Plant Breeding Department, International Rice Research Institute, P.O. Box 933, Manila,
Philippines

272 Rice Breeding and Genetics; Research Priorities and Challenges
and stem borer, while abiotic stresses include drought, salinity, and
submergence.
The genetic variability for some useful traits are either limited in
cultivated rice germplasms or due to changes in insect biotype or disease
race make the cultivars susceptible. In this context, it is important to
broaden the rice gene pool by in introgressing new genes from alien
species to meet the challenges of rice production. The wild species of
Oryza are a rich source of agronomically important genes. However,
several barriers, such as genome incompatibility and chromosome non
homology, limit successful gene transfer (Brar and Khush, 1986; Khush
and Brar, 1992). Recent advances in tissue culture coupled with tools of
plant breeding and biotechnology have the exploitation of alien species
of Oryza in rice improvement possible.
Wild species of the genes Oryza have recently shown several
advantages over other methods of gene transfer for rice improvement.
Alien gene transfer from wild rices in nonhazardous and environment
friendly, which is most important for the safety of human beings.
SPECIATION IN GENUS ORYZA
The genus Oryza comprises two cultivated species and twenty-one wild
species. The Asian cultivated rice, O. sativa {2n ~ 24), is grown
throughout the world while the African cultivated rice, O. glaberrima, is
limited to cultivation in West Africa. The progenitor of O, sativa is the
common wild rice O, rufipogonfO. peretinis that exists in perennial form
as well as annual types such as O. nivara. In a parallel evolutionary
pathway, the progenitor of 0 . glaberrima is O. longistaminata with O.
breviligulata in the annual form (Chang, 1976). The cultivated and related
wild species belonging to the O. sativa complex are easily crossable and
share a common A A genome. Species O. sativa, O. glaberrima, O. nivara,
O. rufipogon, O, breviligulata, O. longistaminata, O. glumaepatula, and O.
meridionalis from the primary gene pool of rice. The wild species
belonging to the O. offidanalis. complex are O. punctata, O. officinalis, O,
rhizomatis, O. eichingeri, O. minuta, O, latifoHa, O. alta, O, grandiglumis, O.
australiensis, and O. hrachyanatha. They have BB, CC, CC, CC, BBCC,
CCDD, CCDD, CCDD, EE, and FF genomes respectively. These species
from the secondary gene pool of rice and are partially homologous or
nonhomo logous to A A genome species, resulting in limited crossover
(Khush, 1977). All the distantly related wild species belonging to the
O. meyeriana complex constitute the tertiary gene pool These species are
O. meyeriana, O. granúlala, O. ridleyi, O. longiglumis and O. schlechteri.
They belong to GG and HHJJ genomes except for O. schlecheri whose

K.K, Jena and G.S, Khush 273
genome is still not known (Aggarwal et al., 1996). The wild species have
either In = 24 or 2n = 48 chromosomes.
USEFUL CHARACTERS OF WILD SPECIES IN ORYZA
The rice O. sativa is grown worldwide under a wide range of
agroclimatic conditions and is affected by several biotic and abiotic
stresses. Even though a resistance source is available in cultivated rice
germplasm/ the resistant varieties are becoming vulnerable to pests and
diseases due to changes in insect biotypes and disease races and as a
result rice productivity has reduced. In order to create genetic variability
and broaden the gene pool of rice, there is a need to look for useful genes
from alien germplasm sources. The wild species of Oryza have a rich
source of genes for resistance to diseases, insects, and several abiotic
stresses (Table 12.1). Even the genes for resistance to. sheath blight,
tungro, and yellow stemborer are available only in some wild Oryza
species and are not present or very limited within the cultivated rice
germplasm. In this context, there is an urgent need to exploit wild
species of Oryza by introgressing agronomically important genes for
broadening the cultivated rice gene pool to increase rice production
(Table 12,2)
Table 12,1
Genome composition and agronomically useful traits of Oryza species
Species Genome Agronomically useful traits’^
O. sativa complex
0 , satwa AA Gültigen
O. nivara AA Resistance to grassy stunt virus and B1
0. rufipogon/perennis AA Elongation ability, resistance to BB, source of CMS
0. glaberrima A®A® Gültigen
0. breviligulata A8A« Resistance to GLH and BB
O, longistaminata ASAS Resistance to BB
0. meriodionalis A»a “' Elongation ability
O. glutmepatula
ASPASP Elongation ability, source of CMS
O. o/A’dnalis complex
0. punctata BB^BBCC Resistance to BPH and ZLH
0. minuta BBCC Resistance to ShB, BB, Bl, BPH, GLH
0 . officinalis CC Resistance to BPH, WBPH, GLH and thrips
0. rhizomatis CC Drought avoidance
Ó. eichingeri CC Resistance to yellow mottle virus, BPH, GLH
O, alta CCDD Resistance to stem borer, high biomass
production
0 , grandiglumis CCDD High biomass production
0 . latifolia CCDD Resistance to BPH, high biomass production
0. australiensis EE Drought tolerance, resistance to BPH
0. brachyanatha FF Resistance to yellow stem borer, leaf folder,
whorl maggot

274 Rice Breeding and Genetics: Research Priorities and Challenges
Table 12,1 Contd,
Species Genome Agronomically useful traits'^
O. meyeriana complex
0 . granulata GG Shade tolerance, adaptation to aerobic soil
O. meyeriana GG Adaptation to aerobic soil
0. ridleyi complex
0. longiglumis HHJJ Resistance to BB, B1
0. ridleyi HHJJ Resistance to stem borer, BB, Bl
0. schlecteri not known —
B1: Blast; B B : bacterial leaf blight; GLH: green leafhopper; ZLH : zigzag leafhopper;
WBPH : white-backed planthopper; ShB ; sheath blight; BPH: brown planthopper.
Table 12.2
Agronomically important genes introgressed from wild Oryza
species into cultivated rice
íKii!
Genes
transferred
to 0 . sativa
Donor
species
Genome
IRRI
accession
number
Grassy stunt '
0 . nivara AA 101508
virus resistance
Bacterial blight 0, longistaminata AA —
resistance 0. officinalis CC 100896
O. minuta BBCC 101141
Blast resistance 0. minuta BBCC 101141
BPH resistance 0. officinalis CC 100896
0. australiensis EE 100882
WBPH resistance 0, officinalis CC 100896
Cytoplasmic male 0. sativa i. spontanea AA —
sterility 0. perennis AA 104823
Yellow stemborer 0. hrachyantha FF 101232
resistance* 0. ridleyi HHJJ 100821
Sheath blight
resistance*
0. minuta BBCC 101141
* Advanced backcross progerües are being produced at IRRI.
ALIEN GENES INTROGRESSED FROM PRIMARY GENE POOL
Several alien genes have been introgressed from the AA genome wild
species into O. sativa. These genes are grassy stunt virus resistance (Gs)
from O. nivara, cytoplasmic male sterility (GMS) from O. sativa
f. spontanea and more recently bacterial leaft blight (BB) resistance from
O. longistaminata. The genes are transferred with no crossability and
recombination barriers.
Introgression of Gene(s) for Resistance to Grassy Stunt Virus
A severe epidemic of grassy stunt virus occurred in rice during the
1970s. The vector for transmission of grassy stunt virus is the brown

K.K. Jena and G.S. Khush 275
planthopper (BPH). The infected rice plants either produce no panicles
or produce small panicles with deformed grains. Of the 6,723 accessions
of cultivated rice and several wild species accessions of Orym screened
at IRRI for resistance, only one accession of O. nivara (Acc. 101508) was
found to be resistant (Ling et al„ 1970). Crosses were made between
improved rice varieties such as IRS, IR20, IR24, and O. nivara and the
resistance gene was transferred into cultivated rice by three backcrosses
with no crossability barrier. Several high-yielding grassy stunt virusresistant
varieties, such as IR28, IR29, IR30, IR32, IR34, and IR36, were
released for cultivation across rice-growing countries. Subsequently,
several other grassy stunt virus-resistant varieties were developed in
different countries.
Introgression of a Gene for Resistance to BB
BB caused by Xanthomonas oryzae pv. oryzae is the one of the most
destructive diseases of rice. A dominant BB resistance gene X«-21 has
been transferred from O. longistaminata into IR24 by backcrossing
(IQiush et aU 1990). This gene has shown a wide spectrum of resistance
to a large number of races of BB (Ikeda et al., 1990).
Introgression of CMS genes from wild Oryza Species
The development of CMS lines with the nuclear genome of rice has been
possible by exploiting the cytoplasm of the wild species O. sativa
f. spontanea. This wild species was discovered in Hainan Island in China
and was identified to have wild abortive (WA) cytoplasm causing male
sterility with abortive pollen. Using this novel source of CMS, it has
become possible to develop high-yielding rice hybrids for commercial
cultivation (Lin and Yuan, 1980). Recently, diversification of CMS from
the WA cytoplasm source to another AA genome wild species became
possible by identifying a new CMS source from O. perennisfO. tufipogon
after making interspecific crosses with 46 accessions of AA genome wild
species (Dalmacio et al, 1995), This new CMS source is IR66707 A with
the nuclear genome of IR64 and is independent of the WA cytoplasm
present in V20A. Another CMS line from O. glumaepatula has also been
identified recently (Dalmacio et al, 1996).
Alien Genes Introgressed from Secondary Gene Pool
Interspecific hybrids between rice and wild species belonging to the
secondary gene pool are difficult to produce. Very low crossability and

276 Rice Breeding and Genetics; Research Priorities and Challenges
degeneration of hybrid embryos during early stages of development are
the major constraints of these crosses. These interspecific hybrids are
male sterile and embryo rescue is a must to develop hybrids and
backcrossing to recurrent O, sativa parent is needed until partially fertile
plants with normal chromosome complement (disomies) having 2n = 24
or monosomic alien addition lines (MAAL) having 2« + 1 = 25
chromosomes are produced (Fig. 12.1), The fertile progenies are selfed
to develop alien gene introgression lines and evaluated for transfer of
desirable agronomic traits.
hji;
Fig. 12,1
An embryo rescue technique used for developing interspecific hybrids and
backcross progenies in rice.
ALIEN GENE INTROGRESSION FROM BB GENOME
Interspecific hybrids have been developed between the autotetraploid
of a japónica cultivar, Nipponbare, and O. punctata {2n = 24; BB). Several

H T »
K.K. Jena and G.S. Khush 277
MAALs and disomie progenies have been produced (Yasui and Iwata,
1991). However^ the progenies have not yet been evaluated for
introgression of useful genes from O. punctata.
Alien Gene Introgression from CC Genome
Interspecific hybrids between cultivated rice and CC genome wild
species have been produced by means of embryo rescue (Jena and Khush,
1984; Fig. 1 2 .2 ), Several introgression lines were produced from this cross
which have iriherited different gènes from O. officinalis (Jena and Khush,
1989,1990). Agronomically important genes for resistance to BPH, whitebacked
planthopper (WBPH) and BB have been transferred into an elite
breeding line of rice (Fig. 12,3). Of the 25 BC2px disomie progenies, 6
segregated for resistance to BPH and 12 segregated for resistance to
WBPH. The recurrent parent IR31917-45-3-2 is susceptible to all
Philippine and Indian biotypes of BPH where as O. officinalis accession
number 100896 is resistant to these biotypes. Several introgression lines
have been produced which are resistant to BPH, biotypes of the
Philippines, India, and Bangladesh (Jena and Khush, 1990).
Fig. 12.2
An interspecific hybrid plant (Fj) produced from a cross between IR31917-45-
3~2 and O. officinalis.

278 Rice Breeding and Genetics: Research Priorities and Challenges
/;
'
H
'
®§i
list r i l f t
a S ; ;
Fig. 12.3
Reaction of several introgression lines derived from O, mliva and O. offiunalis
cross to brown planthopper (BPH). Note several resistant introgression lines
and susceptible lines.
Some of the BPH resistant progenies have been evaluated in replicated
yield trials. Most of the introgression lines had excellent yield
potential and some outyielded the check varieties by a small margin
(Jena and Khush, 1990). Since the selected lines were free from
undesirable traits of wild species^ three BPH resistant lines were released
as varieties for commercial cultivation in the Mekong Delta of Vietnam.
IR54751-2-44-15-24-3 was named as MTL98, IR54751-2-34-10-6-2 as
MTL103, and IR54751-2-41-10-5-1 as MTL 105 (Brar and Khush, 1997).
Some of the resistant lines are also used as donors for resistance in ricebreeding
programs in many countries. Besides resistance to BPH, genes
for resistance to WBPH, BB and some morphological traits such as hull
color, pigmented pericarp, anthocyanin pigmentation of stigma,
apiculus, and leaf sheath inherited from O. officinalis into O. saliva have
been identified (Jena and Khush, unpubl.)
ALIEN GENE INTROGRESSION FROM BBCC GENOME
Interspecific hybrids between O. saliva and the allotetraploid wild
species O. minula {2n = 48) have been produced. Following backcrossing

K.K. Jena and G.S. Khush 279
and embryo rescue^ advanced progenies were developed. The advanced
progenies were evaluated for resistance to BB and blast. One of the two
introgression lines was resistant to Race 6 of BB and another line to Race
P06-6 of blast (Amanta-Bordeos et ah, 1992).
ALIEN GENE INTROGRESSION FROM CCDD GENOME
Interspecific hybrids have been produced between cultivated rice and
different wild species of CCDD genome using the embryo rescue
technique. Following backcrosses and chromosome elimination;
introgression lines inheriting genes for resistance to BPH; WBPH; and
BB from O. latifolia have been developed Qena and Khush, unpubl.)
ALIEN GENE INTROGRESSION FROM EE GENOME
Interspecific hybrids between diploid O. sativa and O. australimsis were
produced by embryo rescue (Jena and Khush, 1984). However, this
particular hybrid did not produce backcross progeny upon repeated
backcrossing of hybrids with the recurrent O. sativa parent. Hence,
interspecific hybrids between colchicine—^induced autotetraploid
cultivated rice and O. australimsis have been produced by embryo
rescue. Following two backcrosses with the recurrent O. sativa parent,
disomic and aneuploid progenies were produced which inherited O.
australiensis traits such as long awn, earliness for flowering, and
resistance to BPH. Evaluation of 600 BC2F4 progenies revealed
introgression of genes for resistance to BPH in four lines and one line
was resistant to Race 6 of BB Qene et ah, 1991; Multani et ah, 1994),
ALIEN GENE INTROGRESSION FROM FF GENOME
The wild species O. brachyantha is resistant to the yellow stem borer,
whorl maggot, and some races of BB. Introgression lines were developed
through production of intraspecific hybrid between O. sativa and the
wild species O. brachyantha followed by backcrossing with the recurrent
parent (IR56). Some introgression lines showed resistance to BB Races 1,
2, 3, 4, and 6 of the Philippines which had been transferred through
limited recombination from O. brachyantha (Brar and Khush, 1997).
However, none of the introgression lines showed resistance to the yellow
stem borer.

280 Rice Breeding and Genetics: Research Pridrities and Challenges
ALIEN GENE INTROGRESSION FROM GG AND HHJJ GENOME
Wild Oryza species belonging to the GG genome are the source of gene
tolerance to water stress or drought and HHJJ genome species are the
source of gene resistance to the stem borer. Herice, efforts have been
made at IRRI to produce hybrids between rice cultvars and these species
(Elloran ei al, 1992). Hybrids have been produced and introgression
lines are being developed for the transfer of agronomically important
traits into rice,
MOLECULAR CHARACTERIZATION OF ALIEN GENES
TRANSFERRED FROM WILD SPECIES INTO
CULTIVATED RICE
I
r
! !
Recent developments in molecular and cellular genetics have made it
possible to map and characterize a few alien genes via linkage to DNA
markers using RFLP (restriction fragment length polymorphism) and
RAPD (random amplified polymorphic DNA) analysis.
Mapping of Xa~21 Gene for BB Resistance
The near isogenic line (NIL) of rice cultivar IR24 containing a gene Xa-21
introgressed from O. longistaminata conferred resistance to all races of
BB. RFLP analysis of this NIL with several DNA markers identified a
marker RG103 located on rice chromosome 11 detected polymorphism
and the marker cosegregated with Xfl-21 gene for BB resistance. All
other DNA markers of chromosome 11 were monomorphic between the
NIL and IR24 (Ronald et al 1992). Further analysis of the NIL with
RAPD markers identified two RAPD markers—^RAPD 818 and RAPD
248—^which cosegregated with resistance locus Xa-21. Physical analysis
of the Xa-21 gene locus revealed a close interrelationship between RAPD
markers (Song et al,1995) and isolated Xa-21 gene by positional cloning;
this gene was subsequently used for genetic transformation of rice for
BB resistance. The transgenic plants carrying the Xa-21 gene expressed a
high level of resistance to the BB pathogen (Wang et al., 1996). This
resistance gene (Xa-21) encodes a putative receptor kinase (Ronald,
1997).
RFLP Analysis of Alien Gene Introgression
Fifty-two introgression lines derived from O. sativa and O. officinalis
crosses have been analyzed withy 188 RFLP markers distributed over
the rice chromosomes (Jena et at,, 1992). Of the 174 informative markers.

ii
K.K. Jena and G.S. Khush 281
only 28 RFLP markers identified the introgression of 0\ officinalis
chromosomal segments in some introgression lines. Introgressed
segments were small and present on 1 1 of the 1 2 rice chromosomes
(Pig, 12.4). Introgression of small segments as observed in this cross
require double crossovers which is in contrast to chromosome pairing
between AA and CC genomes. In most caseS/ O. sativa alleles were
replaced by O. officinalis alleles, indicating reciprocal recombination as
the mechanism of gene transfer between O. officinalis and O. sativa
(Fig. 12.5). A putative DNA marker linked to the BPH resistance gene
derived from O. officinalis was identified but further research is in
progress to confirm it (Jena, unpubl.)
1
T4GZ
. T 636
4 ..7 7
140
636
1430
620
013' 13
6S4
89
744
158 ®
139 6
252 11
6
656
25 10
157
10
171 16
644
152
jf'
104
3M_k.fi
OP
14
944 ' 6
476 0
409 V fi
|329
^11 [J
17
460
227 23
722
746
161
160
122
177
464
i 163
26
666
207
313
403
229
470
[] “
m *
400
213 25
136 6
■■64 , ,
264
123
351_?
■H»
404 %
.10
r f i'* '
424 17
162
172
778 '1
6
244
433
40
■•4
|711
16
156 ^9
170-
173
611
685
|a33
106
27
634 “
432
653
141
125
667
662
1 0 1 1
1 2
363 T 161
8
9S C!]>76
^ 4 T T aM
12
-131 7
3
2
A■•167
'' ■ Ü-17 2«
T7flS
49B
676
30
■ÍÜ
477
03
- 752
^ 1
511
ills
Fig. 12,4
Rice RFLP map showing introgressed segments of O. of^cimlis detected by
RFLP markers. The alien introgressed segments are identified by boxes and
arrows.

282 Rice Breeding and Genetics: Research Priorities and Challenges
Fig. 12.5
Southern hybridization of parents [Lane 2 (O. saliva) and 3 (O, officinalis)] and
one introgression line (Lanes 6) showing replacment of O. saliva allele (->)
with O. officinalis allele (->}. Individual DNA has been cut by Eco RI and
probed with RG 214. Lane 1 is molecular weight marker (A/Hindi III).
Molecular Tagging of BPH Resistance Gene
A gene conferring resistance to three BPH biotypes of the Philippines
was transferred into rice from O. australiensis (Jena et ah, 1991). The
introgression line (IR65482-4-136-2-2) derived from the cross was resistant
to BPH. The of the cross between IR65482-4-136-2-2 and the
susceptible recurrent parent (IR31917-45-3-2) was resistant to BPH, indicating
the dominant nature of the alien gene. This dominant alien gene
for resistance was analyzed, using several RFLP probes of chromosome
12. Of the 14 polymorphic probes analyzed, only RG457 identified
introgression from O. australiensis into O. saliva. Cosegregation analysis
for BPH reaction and RG457 in the F2 population revealed linkage of the
BPH 10(t) gene with a map distance of 3.68 + 1.29 cM (Ishii et ah, 1994).
Molecular Mapping of Blast Resistance Gene
A gene for blast resistance (Pi-9t) was introgressed into rice from the
tetraploid wild species O. minuta. The introgression line carrying the Pfi
9t gene for blast resistance was analyzed with RAPD markers and three
RAPD markers were found to be linked to the Pi-9t gene (Nelson,
unpubL).

FUTURE EXPLOITATION OF ALIEN GERMPLASM
K.K. Jena and G.S. Khush 283
The rapid progress in. introgression of alien genes into cultivated rices
should provide many tools and valuable information to molecular
biologists and plant breeders for crop improvement. The availability of
these genes in the background of O. sativa has already opened up several
new areas of research. Even though resistance genes available in
cultivated rice germplasm have been successfully utilized in 'classical
rice-breeding programs, for crop protection for nearly a century,
limitation of cultivated genetic resources necessitates incorporation of
more and more alien genes so that rice production can be increased by
broadening the genetic base. Recent availability of cloned resistance
genes could provide additional tools for genetic engineering of
improved rice cultivars by transformatiori. QTLs have been identified in
O. rufipogon having wild genes to enhance yield potential after their
transfer into cultivated rice (Xiao et ah, 1996). In spite of the great
potential of genes from wild Oryza species, crossability barriers and
limited recombination are the main constraints limiting interspecific
gene transfer. Future exploitation of alien germplasm must focus on
enhancing genetic recombination between cultivated and wild Oryza
species. Researchers should aim at (1) identifying gene(s) controlling
homologous chromosome pairing in Oryza (2 ) enhancement of alien
gene introgression through tissue culture of wide hybrids and backcross
progenies by promoting recombinational events between cultivated and
wild species genomes, (3) identification of candidate gene(s) through
comparative genome mapping between O. sativa and wild species as
demonstrated by Jena et al (1994). With an increasing number alien
genes to be introgressed into O. sativa, refinement of gene transfer
technology is needed. Such gene transfers will eventually lead to a better
understanding of unknown gene products for disease and insect
resistance in rice and subsequently increased rice productivity.
Acknowledgments
Dr. K.K. Jena is grateful to the World Bank, Robert S. McNamara
Fellowship programs and the Rockefeller Foundation for financial
support for Iris research on alien gene transfer in rice at I.RRI, USA. Both
authors Mr. B.R. Barwale, Chairman, Mahyco Research Foundation for
his encouragement to write this manuscript. We also thank Springer-
Verlag, Germany for allowing us to reproduce Fig. 12.4. We are grateful
to Mr. Bh. V. Sarma who formated the manuscript.

284 Rice Breeding and Genetics: Research Priorities and Challenges
References
i i i
Aggarwal, R,K,, Brar, D.S. and Khush, G.S. 1996. Two new genomes in the Oryzfl complex
identified on the basis of molecular divergence analysis using total genomic DNA
hybridization. Mol. Gen, Genet. 254; 1-12.
Amante-Bordeos, A., Sitch, L.A., Nelson, R., Dalmacio, R.D., Oliva, N.P., Aswindoor, H. and
Leung, H. 1992. Transfer of bacterial blight and blast resistance from the tetraploid wild
rice Oryza minuta to cultivated rice Oryza sativa. Theor. Appl. Genet. 84; 345-357.
Brar, D.S. and Khush, G.S. 1986. Wide hybridization and chromosome manipulation in
cefelas. In; H andbook o f plant cell culture^ Vol. 4: Techniques and applications, D.H. Evans,
W.R. Sharp, P.V. Ammirato (eds,). McMillan Publ Co., New York,TJSA, pp. 221-263.
Brar, D.S. and Khush, G.S. 1997. Alien introgression in rice. Plant M ol. Biol, 35:35-47.
Chang, T.T. 1976. The origin, evolution, cultivation, dissemination and diversification of
Asian and African rices. Euphytica 25: 435-444.
Dalmacio, R.D., Brar, D.S., Ishii, T,, Sitch, L.A., Virmani, S.S. and Khush, G.S. 1995.
Identification and transfer of a new cytoplasmic male sterility source from Oryza perennis
into indica rice (O. sativa). Euphytica 82: 221-225.
Dalmacio, R.D., Brar, D.S., Virmani, S.S. and Khush, G.S. 1996. Male sterile line in rice {Oryza
sativa) developed with O. glum aeputula cytoplasm. IRRN 21(1): 22-23.
Elloran, R., Dalmacio, R.D., Brar, D.S. and Khush, G.S. 1992. Production of backcross
progenies from a Cross of Oryza sativa x O. granúlala. Rice Genet, Nezvslett. 9:39.
Ikeda, R., Khush, G.S. and Tabien, R.E. 1990. A new resistance gene to bacterial blight
derived from O. longistam inata. fpn, J, Breed. 40 (Suppl 1): 280-281.
Ishii, T., Brar, D.S., Multani, D.S. and Khush, G.S. 1994. Molecular tagging of genes for brown
planthopper resistance and earliness introgressed from Oryza australiensis into cultivated
rice, 0. sativa. Genome 37: 217-221.
Jena, K.K. and Khush, G.S. 1984. Embryo rescue of interspecific hybrids and its scope in rice
improvement. Rice Genet. Newslett. 1 ; 133-134.
Jena, K.K. and Khush, G.S, 1989. Monosomie alien addition lines of rice: production,
morphology, cytology and breeding behaviour. Genome 32:449-455.
Jena, K.K. and Khush, G.S. 1990. Introgression of genes from O ryza officinalis Wall ex Watt to
cultivated rice, O. sativa L. Theor. Appl. Genet. 80: 737-745,
Jena, K.K., Multani, D.S. and Khush, G.S. 1991. Monosomie alien addition lines of Oryza
australiensis a n d alien gene transfer. R ice Gen. II: 728.
Jena, K.K., Khush, G.S. and Kochert, G. 1992. RFLP analysis of rice (Oryza sativa L.)
introgression lines. Theor. Appl. Gen. 84: 608-616.
Jena, K.K,, Khush, G.S. and Kochert, G. 1994. Comparative RFLP mapping of a wild rice,
O ryza officinalis and cultivated rice, O. sativa. Genome 37:382-^89,
Khush, G.S. 1997. Origin, dispersal, cultivation and variation of rice. Plant M ol. Biol 35:25-
34.
Khush, G.S. and Brar; D.S. 1992. Overcoming the barriers in hybridization. Theor. Apl. Gen.
(Monograph No. 16) pp. 47-61.
Khush, G.S., Bacalangco, E. and Ogawa, T. 1990. A new gene for resistance to bacterial blight
from O. longistam inata. R ice Gen. Newslett. 7:121-122,
Lin, S.C. and Yuan, L.P. 1980.. Hybrid rice breeding in China. In: Innovative approaches to Rice
Im provem ent. IRRI, and Manila, Philippines, pp 35-51.
Ling, K.C., Aguiero, V.M. and Lee, S.H. 1970. A mass screening method for testing resistance
to grassy stunt disease of rice. Plant Dis. Rep. 56:565-569,

K.K. Jena and G.S. Khush 285
Multani, D.S,. Jena, K.K., Brar, D.S,. délos Reyes, B.C., Angeles, E.R. and Khush, G.S. 1994.
Development of monosomic alien addition lines and introgression of genes from Oryza
australiensis Domin. to cultivated rice O. sativa. Theor. Appl. Genet. 88:102-109.
Ronald, P.C. 1997. The molecular basis of disease resistance in rice. Plant M ol. Biol. 35:179-
186.
Ronald, P.C., Albano, B., Tabien, R-, Abenes, L., Wu, K., McCouch, S. and Tanksley, S.D. 1992.
Genetic and physical analysis of rice bacterial blight resistance locus Xfl-21. M ol. Gen.
Genet. 236:113-120.
Song, W.Y., Wang, G.L., Chen, L.L., Kim, H.S., Pi, Y.L., Hols ten, T., Gardner, J., Wang, B.,
Zhai, W. X,. Zhu, L.H., Fraquet, C. and Ronald, P. 1995. A receptor kinase like protein
encoded by the rice disease resistance gene, Xii-21. Science 270:1804-1806.
Wang, G.L., Holsten, T.E., Song, W.Y., Wang, H.P. and Ronald, P.C. 1995. Construction of a
rice bacterial artificial chromosome library and identification of clones linked to Xa-21
disease resistance locus. The P la n t}. 7: 525-533.
Xiao, J., Grandillo, S., Ahn, S.N., McCouch, S.R., Tanksley, S.D. and Yuan, L.P. 1996. Genes
from wild rice improve yield Nature 384: 223-224.
Yasui, H. and Iwata, N. 1991. Production of monosomic alien addition lines of Oryza sativa
having a single O. punctata chromosome. Rice Genetics II: 147-155.
i

13
Cytogenetics of Rice
R,J. Singh^ and G.S. Khush^
INTRODUCTION
Rice and wheat are the world's most important food crops. World rice
production in 1996 was 373.26 million metric tons (Mmt) while wheat
accounted for 609.57 Mmt (FAO, 1996). Rice is the principal staple food
source for more than half mankind. However, about 92% of all rice is
produced and consumed in Asia (Khush, 1975, 1997).
Cultivated rice belongs to the genus Oryza L,, subfamily Oryzoideae,
in the family Poaceae {Gramineae). The genus Oryza is extremely variable
and distributed in tropical and temperate regions of the world
(Vaughan, 1994). Rice is cultivated between 36ES to 55EN and grows
from sea level to an altitude of 2^500 m or even higher (Khush and Singh,
1991; Khush, 1997).
During the past two decades our knowledge of rice cytogenetics has
enhanced at a rapid pace. Rice has become a model monocotyledonous
cereal plant for classical, biochemical, and molecular genetic studies
(Izawa and Shimamoto, 1996). The taxonomy of the genus Oryza is well
defined (Vaughan, 1994) and genomic relationships among species have
been established by classical, cytogenetic, and molecular methods (Nezu
et at, 1960; Ogawa and Katayama, 1973, 1974; Katayama and Ogawa,
1974; Katayama and Onizuka, 1978; Katayama et at, 1977; Katayama,
1977,1995, 1997; Nayar, 1973; Oka, 1964; Wang et al, 1992; Aggarwal et
^ Department of Crop Sciences, University of Illinois, Urbana, IL 61801
International Rice Research Institute, Los Baños, Philippines

\ ]
288 Rice Breeding and Genetics; Research Priorities and Challenges
at., 1997). Rice is a diploid {2n = 24), has a small genome size (2.02 pg)
among major crops (Bennett and Leitch^ 1997), and has established
cytogenetic stocks such as primary trisomics^ secondary trisomics^
telotrisomics^ translocations^ and monosomic alien addition lines
(MAALs). The cytogenetic stocks have been employed to develop
cytologicab classical^ and molecular linkage maps of rice (Khush el ah,
1984; Iwata and Omura^ 1984; McCouch et ah, 1988; Singh et ah, 1996a,
1996b; Yu et ah, 1995; Khush et ah, 1996; Chen et ah, 1997). The production
of a series of MAALs from wide crosses has enabled the
introgression of economically useful alien genes into cultivated rice
(Jena and Khush, 1989; Jena et ah, 1992; Multani et ah, 1994).
BIOSYSTEMATICS
; I
The genus Oryza belongs to the subfamily Oryzoideae, tribe Oryzeae in the
family Poaceae {Gramineae). Vaughan (1994) described 23 distinct species
in the genus and grouped them into four species complexes. These are:
O. saliva complex, O, officinalis complex, O. meyeriana complex, and O,
ridleyi complex. O. schlechteri does not belong to any of these complexes.
O, schlechteri was thought to be extinct but was recently recollected by
Vaughan (1994) from Papua New Guinea. The genomes of O. schlechteri
are not known.
1. O. sativa complex; This complex includes two cultigens, O. sativa
(indica and japónica rices) and O. galberrima (African cultivated rice),
and six wild species. O. sativa is grown worldwide while O. glaberrima is
limited to tropical West Africa. All species have 2n = 24 chromosomes
and an A A genome (Table 13.1).
2. O. officinalis complex: This is the largest complex in the genus. It is
composed of ten species. Five species are diploid (2n = 24), four
tetraploid (2n - 48), and one (O. punctata) contains both diploid and
tetraploid cytotypes (Table 13.1). Tetraploid species are allotetraploids.
The species of this complex are distributed in Asia, Africa, and Latin
America. All species from South and Central America are tetraploid
with CCDD genomes and tetraploid species from Asia and Africa carry
BBCC genomes (Table 13.1).
3. O. meyeriana complex: The O. meyeriana complex consists of O.
granúlala and 0 . meyeriana. Both species have 2n = 14 chromosomes. Of
the two, O. granúlala is widely distributed and thrives in South Asia,
Southeast Asia, and south western China while O. meyeriana is
distributed in South east Asia. The O. meyeriana complex also includes
the recently named taxon Oryza indandamanica Ellis from Rutland Island
in the Andamans. However, Vaughan (1994) considered A Oryza
indandamanica @ an isolated population of O. granúlala. This observation

R.J. Singh and G.S, Khush 289
Table 13.1
Chromosome number, genomic composition and potential
useful traits of Oryza species (Khush, 1997).
Species 2n Genome Distribution Useful or potentially useful traits*
(1) (2) (3) (4) (5)
l.O. sativa complex
0 . satim L. 24 AA Worldwide Gültigen
0 . nivara Sharma 24 AA Tropical and sub- Resistance to grassy stunt
etShastry tropical Asia virus, blast, drought
avoidance
0. rufipogon Griff. 24 AA Tropical and sub- Elongation ability, resistance to
tropical Asia BB, source of CMS
0 . breviligutata 24 A W Africa Resistance to GLH, BB, drought
A. Chev. et Roehr. avoidance
0 . glaberrima Steud. 24 West Africa Gültigen
0. longistaminata 24 ASAS Africa Resistance to BB, drought
A. Chev. et Roehr. avoidance
0 , meridimalis Ng 24 A^^A"' Tropical Australia Elongation ability, drought
avoidance
0. glumaepatula 24 ASP ASP South and Central Elongation ability, source
Steud.
America
of CMS
II. O. offidnalis
complex
0 . punctata 24 BB Africa Resistance to BPH
Kotschy ex Steud. 48 BBCC Zigzag leafhopper
0 . minuta J.S. Presl. 48 BBCC Philippine and Resistance to sheath blight.
exC.B.Presi.
Papua New Guinea BB, BPH, GLH
0 . officinalis 24 CC Tropial and sub- Resistance to thrips,
Wall ex Watt tropical Asia BPH,GLH,WBPH
0 , rhizomaiis 24 CC Sri Lanka brought avoidance,
Vaughan
rhizomatous
O. eichingeri 24 CC South Asia and Resistance to yellow mottle
A. Peter East Africa virus,
BPH,WBPH,GLH
0 . latifolia Desv. 48 CCDD South and Central Resistance to BPH, high
America
biomass production
O. alta Swollen 48 CCDD South and Central Resistance to striped stem
America
borer, high biomass production
O. grandiglumis 48 CCDD South and Cenh'al I ligh biomass production
(Doell) Prod.
America
O, australiensis 24 EE Tropical Australia Drought avoidance, resistance
Domin
to BPH
0 . brachyantha 24 PF Africa Resistance to yellow stem
A. Chev. et Roehr. borer.
Leaf folder, whorl maggot,
tolerance to laterite soil
in. O.meyetiana
complex
O. granulata 24 GG South and Shade tolerance.
Ness et Southeast adaptation to
Arn.ex Walt Asia aerobic soil
0 . meyeriana 24 GG Southeast Asia Shade tolerance, adaptation to
(Zoll, et Mor. ex
Steud.) Baill.
aerobic soil
(Confd.)

290 Rice Breeding and Genetics: Research Priorities and Challenges
Table 13.1 Contd.
Species 2n Genome Distribution Useful or potentially useful traits*
(1) (2) (3) (4) (5)
IV. O. ridleyi
complex
0 . longiglumis 48 HHJJ Irian Jaya, Resistance to blast, BB
Jansen
Indonesia and
Papua New Guinea
O. ridleyi Hook. HHJJ South Asia Resistance to stem borer, whorl -
f.48 maggot, blast, BB
Genome not known
O. schlechteri 48 Not Papua New Stoloniferous
Pilger known Guinea
* EPH ; brown planthopper; GLH ; green leafhopper; WBPH : white-backed planthopper;
BB : bacterial blight; CMS : cytoplasmic male sterility,
If MH
is supported by Aggarwal et al. (1997) from molecular studies. They
demonstrated that O, meyeriana, O, granulata and O. indandamanica have
a similar genome and have been assigned the genome symbol GG.
4. Oryza ridleyi complex: This complex contains O. longiglumis an
O, ridleyi. These two species are allotetraploid. O. longiglumis is found in
Irian Jaya, Indonesia and Papua New Guinea while O. ridleyi is
distributed in South Asia (Table 13.1). Taxonomically, the two species
are almost similar and clearly distinct from other species of the genus.
Based on total genpmic DNA hybridization, Aggarwal et al. (1997)
proposed a new genome symbol, HHJJ, for species of the O. ridleyi
complex.
Phylogenetic relationships among diploid and tetraploid species of
the genus Oryza, and genome designations based on classical taxonomy
and cytogenetics are being verified by isozyme banding patterns
(Second, 1982), nuclear RFLPs (Wang et al, 1992), total genomic DNA
hybridization (Aggarwal et ah, 1997), and simple sequence repeats (Chen
et ah, 1997).
ORIGIN OF CULTIVATED RICES
The genus Oryza originated in the ancient Gondwanaland supercontinent
and consequent to continental drift became widely distributed in
the humid tropics of Asia, Africa, South America, and Oceania (Chang,
1976; 1985). Wild annual O. nivara, derived from O. rufipogon, is
considered the progenitor of O. sativa. O. glaherrima was domesticated
from the wild armual. O. breviligulata in West Africa and is referred to as
the African cultivated rice.
It is evident that cultivated rice was derived through an evolutionary
process following the order wild perennial {wild annual}
cultivated annual (Harlan, 1965). Figure 13.1 shows the evolutionary

R.J. Singh and G.S. Khush 291
.Gondwana land
Common ancestor
South and Southeast Asia
Wild perennial
Wild annual
O. aifipogon
i
0. nivara
Gültigen 0. sativa O. satíva
indica japónica
temperate
tropical
Fig. 13.1
Diagrammatic representation of the spéciation of the two cultivated rices
(from Khush, 1997).
pathiyays of the two taïionomically distinct cultivated species of rice.
Indica-Japonica hybrids are partially sterile due to genic imbalance
(Bouharmont et al, 1985). The progenitor of Asiatic rices is O. nivara and
the Fi hybrids between O. nivara and O. sativa show essentially normal
chromosome pairing and seed fertility (Dolores et al, 1979).
O. glaberrima and its progenitor O. breviligulata are less diverse than
their Asian counterparts (Chang, 1976). Based on isozyme
polymorphism. Second (1982) proposed that O. glaberrima was
domesticated independent of O. sativa. Asiatic and African cultivated
rices and their wild ancestors carry similar genomes (Table 13,1).
KARYOMORPHOLOGY
Numerous attempts have been made to characterize individual
chromosoines of rice and to prepare an idiogram by using somatic
metaphase chromosomes. Hu (1964) measured chromosomes from 2 1
mitotic metaphase haploid cells. Chromosome length ranged from 4.32
m (Chromosome 1) to 1.79 (¡tm (Chromosome 12). Somatic mitotic
metaphase chromosomes of rice are not well defined and lack

292 Rice Breeding and Genetics: Research Priorities and Challenges
I
M
morphologically distinguishing landmarks. It is not possible to identify
all members of the chromosome complément of rice except nucleolus
organizer chromosomes.
Shastry et al. (1960) examined the entire chromosome complement
of Japónica rice cv. Norin- 6 at the pachynema. They were able to
distinguish 1 2 pachytene bivalents based on total length, arm ratio, and
presence or absence of dark-staining heterochormatic chrqmomeres.
Chromosomes were arranged in the descending order of their length,
with the longest (79.0 (pm) as chromosome 1 and the shortest (18.0

R.J. Singh jand G.S. Khush 293
primary trisomics {2n = 2x + 1 - 25) were helpful in delimiting the
position of centromeres, A global uniform chromosome numbering
system for rice was accepted by the Rice Genetics Cooperative during
the Second International Rice Genetics Symposium held at the
International Rice Research Institute (IRRI), Las Banos^, Philippines in
May 1990 (Khush and Singh, 1991).
Based on cytological observations and a physical map of the rDNA
loci, the chromosomes of O. sätiva and O, glaherrima and almost identical
(Ohmido, 1995; Ohmido and Fukui, 1995). Fukui et al. (1994) reported 2
rDNA loci (NOR) in tropical rice and one rDNA locus in rice from the
temperate regions.
ANEUPLOIDY
Primary Trisomics
Rice primary trisomics contain a normal chromosome complement plus
an extra complete chromosome (2n = 2x + 1 = 25). Primary trisomics
have been used extensively for determination of gene-linkage group
relationships in several plant species (Burnham, 1962; Khush, 1973;
Singh, 1993).
,(i) O r ig in o p p r im a r y t r is o m ic s
Several attempts have been made to generate primary trisomics in rice
from the progenies of autotriploids (2ti ~3x = 36) (Hu, 1968; Watanabe ef
al., 1969; Khush et aL, 1984). However, a complete set of all possible
primary trisomics was produced only by Khush et al, (1984) in the Indica
rice cv. IR 36 and by Iwata and Omura (1984) in the Japónica rice cv.
Nipponbare. Of the 92 seeds harvested from an autotriploid plant,
Khush et al. (1984) recovered 72 plants. Twenty plants were primary
trisomics {2n - 25), plants double trisomics (2n = 26) and the remaining
plants segregated for 2n = 24 (2 plants), 2n = 27 (14 plants), 2n = 28 ( 8
plants), and 2n 29 (3 plants) chromosomes. Thus, plants with 2n = 25
and 26 predominated in the progenies of rice autotriploid and the
maximum number of extra chromosomes tolerated by the rice is six.
Therefore, the tolerance limited for the extra chromosomes in rice is
narrow as very few plants with more than four extra chromosomes were
produced (Khush et al, 1984). This may be due to the fact that male and
female spores or zygotes or embryos with more than three extra
chromosomes abort in the progenies of autotriploid of the diploid
species because duplication of extra genetic material causes genetic and
physiological imbalance. It has been observed that the initial phase of

294 Rice Breeding and Genetics: Research Priorities and Challenges
ii ' '■
seed development is normal in autotriploids^r but the endosperm shrivels
after a week, resulting in the death of the embryos. The failure in
endosperm development is caused by an extremely unbalanced
chromosome number and this is the most likely explanation for the high
frequency of occurrence plants with 2s: + 1, + 2, 2x + 3, and 2x + 4
chromosomes in the progenies of autotriploids of diploid species and
lack of plants with a higher chromosome number (Singh 1993).
Ishiki (1991) isolated 8 of the possible 12 primary trisomics in O,
glaberrima from the progenies of an artificially synthesized autotriploid.
Primary trisomics of O. glaberrima were morphologically similar to the
primary trisomics of O. sativa,
(II) M o r p h o l o g ic a l id e n t if ic a t io n o f p r im a r y t r is o m ic s
1^ »
Primary trisomics in indica and japónica rices differ from their normal
diploid sibs and also from each other in several distinctive
iporphological traits (Iwata et ah, 1970; Khush et ah, 1984). The
distinguishing morphological features of each rice trisomics established
in the Indica rice cv. IR36 are as follows:
Triplo 1 Grassy, short height; narrow and thin pale green leaves; late
flowering; narrow and triangular grains; low fertility.
Triplo 2 Short height, few tillers; short dark green thick and twisted
leaves; short panicles; empty glumes; depressed palea; short
anthers and reduced filament; highly self-sterile but produces
abundant seed when pollinated with a diploid.
Triplo 3 Short height; slow growth, reduced tiller numbers; short,
dark green and thick leathery leaves; incompletely exerted
panicles; late flowering; highly male and female sterile.
Triplo 4 Tall height (taller than all of the other primary trisomics and
also diploid sibs); long and droppy light green leaves; long
ligule; lax panicles; high seed fertility.
Triplo 5 Short height; short twisted leaves with fine hairs; short ligule;
short compact panicles; high seed fertility.
Triplo 6 Short height; thick sernirolled dark green leaves; long ligule;
early flowering; lax and awned panicles; partially sterile.
Triplo 7 Narrow, dark green and sernirolled leaves with short ligule;
incompletely exserted and somewhat lax panicles; long tip
awned grains.
Triplo 8 Narrow, dark green, rolled leaves; short height, dense and
fully exserted panicles; short and bold grains; partial seed
fertility.
Triplo 9 Spreading growth habit; dark green leaves; thick stems;
largest panicles among the primary trisomics and the highest
1 0 0 -grain weight.

R J. Singh and G.S. Khush 295
Triplo 10 Fine foliage and stems at flowering stage; erect leaves with
hairy auricles; slender grains; completely fertile;
distinguishable only after flowering.
Triplo 11 Morphologically similar to a diploid, but at booting stage
. plants are slightly golden colored; hull gold color.
Triplo 12 Bushy; many tillers; pale green; lax panicle; degenerated
florets at tip of panicles; long grains; self-fertile.
(ill) C y t o l o g ic a l id e n t if ic a t io n o f p r im a r y t r is o m ic s
Primary trisomics of rice were identified and designated based on
pachytene chromosome analysis (Khush et ah, 1984) and mitotic
metaphase chromosome karyotype (Kurata, 1986). The three homologous
chromosomes in primary trisomics compete to pair with one
another, but only two by two pairing is observed. The third one attempts
to associate with the paired homologues in a random manner and may
form a loose trivalent configuration or pairs with itself (Fig. 13.2). In rice,
the primary trisomic with chromosome 1 in triplicate was called triplo 1 ,
that having an extra chromosome 2 was called triplo 2, and so on (Table
13.2).
Table 13.2
Present status of aneuploid stocks in rice (Khush et al,
1984; Singh et al., 1996b)
Primary trisomics Secondary trisomics Telotrisomics
Triplo 1 2n + IS = IS; 2tt + lL = lL 2n + mlS
Triplo 2 2m+ 2S = 2S; 2n + 2L = 2L 2n+ = 2L
Triplo 3
2n + = 3L
Triplo 4
2ji + 4S = 4S
Triplo 5 2w + 5S s 5S 2n + s 5L
Triplo 6 2« + 6S = 6S; 2« + 6L = 6L
Triplo? 2n + 7S = 7S; 2« +7L^7L
Triplo 8 2n+8L = 8L 2n + s8S
Triplo 9 2w + 9L = 9L 2n + = 9S
Triplo 10
2n + = 10S
Triplo 11 2n -i l l S s l l S ; 2n + 11L = 11L
Triplo 12
2n + 12S s 12S
(IV) T r a n s m is s io n o f t h e e x t r a c h r o m o s o m e
Theoretically, about 50% primary trisomic plants are expected in the
progenies of primary trisomics when used as a female. However, such a
proportion is rarely observed (Khush, 1973; Singh, 1993). The
transmission rates of extra chromosomes of the primary trisomics
through the female in rice are high (Khush et al, 1984). They range from
15.5% (triple 1) to 43.9% (triple 4). Male transmission of the extra
chromosome was recorded in 7 of 12 primary trisomics that ranged

296 Rice Breeding and Genetics: Research Priorities and Challenges
from 0.5% (triplo 4) to 27.3% (triplo 9). The longer chromosomes of the
rice complement did not transmit through the mate gametes. Perhaps
they cause greater physiological and genetic imbalance on the male side.
ThuS/ the behaviour of the primary trisomics of rice is similar to that
observed in several diploid species such as tomato^, barley, and maize
(Khush, 1973; Singh, 1993).
(V) P r im a r y t r is o m ic s in r ic e l in k a g e m a p p in g
Primary trisomics are powerful cytogenetic tools for locating a gene on a
particular chromosome, verifying the independence of linkage groups,
and for associating the genetic linkage groups with individual
chromosomes (Burnham, 1962; Hermsen, 1970; Khush, 1973; Singh,
1993). When a primary trisomic is used to locate a gene on a particular
chromosome, the genetic ratios are modified from 3:1 (F2) or 1:1
(BCi).The ratios depend on the genotypes of the F| primary trisomic
plants [whether duplex (AAa) or simplex (Aaa)], on the type of
segregation (random chromosome or random chromatid), and on the
female transmission rate of the extra chromosome (50% or 33.3%).
The expected phenotypic frequencies in an F2 population, assuming
50% female transmission of the extra chromosome, and duplex genotype
of the F^, would be 17:1 [9:0 (2x + 1):: 8:1 (2^:)] instead of 3:1. however,
the expected 50% female transmission of n + 1 gametes is not usually
observed. If we assume the female transmission of the extra
chromosome as 33.3%, the overall phenotypic ratio is modified from
17:1 to 12.5:1.
Associations between chromosomes and linkage groups through
primary trisomic analysis is in rice were established by Iwata and
Omura (1975, 1976; 1984), Iwata et al. (1984), and Khush et at (1984).
Iwata and Omura (1976) associated three linkage groups of Nagao and
Takahashi (1963) with one chromosome by using primary trisomics.
Khush et al. (1984) examined 120 combinations involving 22 genes and
12 primary trisomics. They located marker genes for each of the 12
chromosomes. Subsequently, several morphological rharkers (Librojo
and Khush, 1986; Iwata et al., 1984; Sanchez and Khush, 1994), isozyme
markers (Ranjhan et al., 1988; Wu et al, 1988), and RFLP markers
(McCouch et al, 1988) were assigned to saturate the linkage maps of rice
(Fig. 13.3).
Kamisugi et al. (1994) physically localized 5S rDNA locus on
chromosome 11 of Japónica rice cv. > Nipponbare=. The location differs
in Indica rice. Wang et al. (1995b) located repetitive DNA sequence in the
heterochromatic region of the long arm of chromosome 5 by fluorescence
in situ hybridization.

R.J. Singh and G.S. Khush 297
Secondaxy Trisomies and Telotrisomics
In secondary trisomies, the extra chromosome is an isochrornosome
(both arms are homologous, e.g. secondary chromosome). In the
telotrisomics, the extra chromosome is a telocentric chromosome. A
telocentric chromosome consists of a centromere and one complete arm
of a normal chromosome.
(I) O r ig in o f s e c o n d a r y a n d t e l o t r is o m ic s
Secondary and telotrisomics are produced as a result of misdivision of
the univalent. The probability of misdivision of univalent in primary
trisomies in higher because many sporocytes contain an extra chromosome
as a univalent. The frequency of secondary and telotrisomics in the
progenies of primary trisomies of rice is shown in Table 13.3. Singh et al.
(1996a) isolated 15 secondary trisomies and 7 telotrisomics. Secondary
trisomies in rice for both arms of chromosomes 1, 2 , 6 , 7, and 11 and for
only one arm of chromosomes 4, 5, 8 , 9, and 1 2 were isolated.
Telotrisomics in rice for 2L, 3L, 5L, IS, 8 S, 9S, and lOS were identified
(Table 13.3.).
Table 13.3
Frequency of secondary and telotrisomics in the progenies of primary
trisomies (Singh et ai, 1996 a).
Trisomic
Total
plants
grown
Secondary trisomic
Short arm
(No.)
Long arm
(No.)
Short arm
■ (No.)
Telotrisomic
Long arm
(No.)
Frequency
Triplo 1® 1 1
lb
0
Triplo 2 1632 2 1*’ 0 1 0.18
Triplo 3“ 0 0 0 1
Triplo 4 1812 1 0 0 0 0.05 "
Triplo 5 1536 1 0 0 1 0.13
Triplo 6 2112 3 1 0 . 0 0.19
Triplo 7 3300 2 1 0 0 0.09
Triplo 8 1608 0 3 1 0 0.25
Triplo 9 2127 0 3 1 0 0.19*
Triplo 10 1776 0 0 1 0 0.06 ’
Triplo 11 600 1 1 0 1 0.50
Triplo 12 1632 2 0 0 0 0.12
** Triple 1 and triplo 3 are highly sterile and large populations could not be grown. 2n + IS ■
IS and 2n + IL •IL and ■3L were selected from progenies of primary trisomics before
conscientious efforts were made to isolate secondary and telotrisomics
Telotrisomic 2n + ■IS was isolated from the progeny of 2n + IS •IS and secondary trisomic
2n + 2L •2L was isolated from the progeny of 2« + ■2L.
(%)

298 Rice Breeding and Genetics: Research Priorities and Challenges
(II) M o r p h o l o g ic a l id e n t if ic a t io n o f s e c o n d a r y a n d t e l o t r is o m ic s
In general/ secondary trisomics expressed slower vegetative growth rate
and lower seed fertility than the corresponding primary trisomics.
However/ they resembled their counterpart primary trisomics in
morphological traits (Singh et al., 1996a). Some of the morphological
traits of primaries were exaggerated in secondary trisomicS/ particularly
those for long arm. Secondary trisomics for short arms were generally
vigorous and were fertile. Secondary trisomics for IS and 2S were
exceptions. They were weak and slow in vegetative growth and pollen
fertility ranged from 11% for the secondary trisomic for IS to 16.8% for
the secondary trisomic for 2S. Secondary trisomics for 4S and IIS were
like diploid sibs.
Telotrisomics were vigorous and fertile compared to their counterpart
primaries and secondaries (Singh et al., 1996a). Telotrisomics for
5L, 8 S/ 9S, and lOS exhibited normal pollen and seed fertility while
Telotrisomics for 2L and 3L showed partial pollen and seed fertility.
Pollen fertility in telotrisomic 2L was 54% and seed set after selfing
ranged from 15”20%. Telotrisomic 3L had 78% pollen fertility and 30-
40% seed set upon selfing.
(III) C y t o l o g ic a l id e n t if ic a t io n o f s e c o n d a r y a n d t e l o t r is o m ic s
Secondary trisomics were first identified cytologically at diakinesis. The
occurrence of a ring configuration suggests that the extra chromosome is
an isochromosome. The frequency of ring trivalent ranged from 0.5%
(2 n + 4S = 4S) to 25.6% {2x + 8 L = 8 L). Usually/ secondary trisomics for
the long arms showed higher frequency of ring trivalents than those for
secondary trisomics for the short arms (Singh et al., 1996a). In secondary
trisomics, the isochromosome can pair internally and remain as a
univalent or pair with the homologous arms of the two normal
chromosomes to form a Y-shaped trivalent. Pachytene trivalent
configurations of secondary trisomics facilitated the precise location of
centromeres. The centromere position of ten chromosomes was verified
and chromosome 6 was found to be more metacentric and chromosome
12 was observed to be submetacentric. A revision of pachytene idiogram
for chromosomes 6 and 12 was suggested (Singh et ah, 1996a).
In telotrisomics, sporocytes with 1211 + II predominated and ranged
from 49% {2n + s 2L) to 87% (2n + ^ 9S). The frequency of llll + IIII
ranged from 13% (2n + s 9S) to 45.9% {2n + = 2L). The telotrisomics for
the long arm formed trivalents at a higher frequency than telotrisomics
for the short arms. Telocentric chromosomes were more difficult to
identify at the pachynema than isochromosomes (Singh et al., 1996a),

RJ. Singh and G.S. IChush 299
(iv) T r a n s m is s io n r a t e s o f t h e e x t r a c h r o m o s o m e in s e c o n d a r y a n d
TELOTRISOMICS
The female transmission rates of extra isochromosomes in selfed
progenies of secondary trisomics ranged from 8.1% (2n + IS = IS) to
47.3% (2« + 4S = 4S). Related primary trisomics appeared in the
progenies of secondary trisomics. Their frequency ranged from 1.4% (2«
+ IS E IS) to 2,07% {2n + 8 L = 8 L). Male transmission of isochromosomes
was recorded only in 2« + 4S = 4S (Singh et at, 1996a),
Female transmission of the extra telocentric chromosome in selfed
progenies ranged from 28.6% (2« + e 2L) to 47.5% (2n + = 9S), higher
than the transmission rates of isochromosomes (Singh et al., 1996a), As
expected, the transmission rates of telocentric chromosomes for the
short arms were higher than those for the long arms. Male transmission
of the telocentric = 8 S was 12% and that of s 9S was 20%.
(v) Lo c a t io n o f t h e g e n e s o n c h r o m o s o m e a r m s
Secondary trisomics can be used to locate genes on a particular
chromosome arm in much the same way as the primary trisomics. If a
gene is located in the extra chromosome arm, a ratio of 3:1 :: all :0 is
observed in F2 for diploid and secondary trisomic fractions. Thus no
recessive homozygous plants are obtained in the secondary trisomic
fraction. In the progenies of telotrisomics on the other hand, teiotrisomic
plants with recessive phenotype can be obtained.
Segregation of 43 marker genes belonging to 11 linkage groups of
rice was studied in the progenies of secondary and telotrisomics (Singh
et al., 1996a). The same marker gene was crossed with the secondary
trisomics for both arms to determine arm location of the gene. For
example, four morphological markers (z-1 , v-A, la, z2 ) were used in
genetic studies with the secondary trisomics for chromosome IIS and
IIL. The F2 segregation ratios conclusively demonstrated that marker
genes v~4, la, and Z2 are located on IIL and z-1 on IIS (Table 13.4).
Segregation data of secondary and telotrisomics facilitated location
of genes on specific chromosome arms and helped determine the
centromere position on 8 linkage groups and orientation of 1 0 linkage
groups of rice. The relationships between (from left to right) the
pachytene idiogram or rice, molecular linkage map, and classical map
'are shown in Fig, 13.3. (Khush et al., 1996). Singh el al. (1996b) assigned
more than 170 RFLP markers to a specific arm of a chromosome by using
secondary and telotrisomics through gene dosage analysis. The
orientation of seven linkage groups was reserved to fit the > short arm
on top-convention.

300 Rice Breeding and Genetics: Research Priorities and Challenges

1 . 1
¡■!
R J. Singh and G.S. Khush 301
Chromosomal Interchanges
A series of reciprocal translocation lines have been produced and
utilized for genetic and linkage studies in rice (Nishimura, 1961;
Yoshimura et al., 1982; Sato and Shinjyo, 1995). Chung and Wu (1994)
identified chromosomes involved in 1 1 of Nishimura=s translocation
lines through pachytene analysis. Nonomura et al. (1997) used
chromosomal interchanges of rice to determine the orientation of RFLP
linkage groups on pachytene chromosomes and the assignment of 113
RFLP markers and 4 cloned rice genes to chromosome arms.
I !
Monosomies, Aneuhaploids, and Tetrasomies
In a monosomie plant, one complete chromosome is missing from the
normal chromosome complement. Seshu and Venkataswamy (1958)
isolated a monosomie plant (2n = 2x - 1 = 23) in rice from the progenies
of indica x japónica hybrids. The monosomie plant was weak and
progenies (45 plants) from this plant were normal and diploid. Wang et
al. (1991) produced monosomies by using gamma-ray irradiated pollen.
In addition, they identified plants with 2n = 23 + If (fragment), + 2f, + 3f.
They used induced deficiencies to locate genes at a particular region of
the rice chromosome.
Aneuhaploid plants contain an extra chromosome in addition to
haploid chromosome complement. Wang and Iwata (1991, 1995)
produced eight aneuhaploids from the anther culture of primary
trisomies of rice cv. > Nipponbare=. These aneuhaploids were for
chromosomes 4, 5, 6 , 8 , 9, 10,11, and 1 2 . Morphologically, aneuhaploid
plants resembled their counterpart primary trisomies. At metaphase I,
chromosome pairing in monosomie plants was mostly III + III or 131.
A tetrasomie individual contains a normal chromosome
complement plus a pair of homologous chromosomes. Wang et al.
(1995a) isolated eight tetrasomies (2n = 26) for chromosomes 4, 5, 6 , 7, 8 ,
9, 10, and 12 from the anther culture of primary trisomies of rice. The
morphological features of’ these tetrasomies were similar to their
corresponding primary trisomies. The extra chromosomes caused a
greater genetic imbalance in haploids than those observed in diploids:
More than 80% of the sporocytes in tetrasomies showed llll + 1 IV
Pollen and seed fertility in tetrasomies is low and thus their use in
cytogenetic studies in limited. RFLP analysis verified the morphological
and cytological identity of the aneuhaploids and tetrasomies (Wang et
al., 1995b).

----RGlQJl
JtQyii '
RG944
SPf3|
-;
(i>
5'
aq £Us
o
fD
3 OS
■RZ3B4,
. RZ58?‘
. R7;^9a
-■“ RZ57G'
■RZIR
RZ403
. ftZT4&
- ftI575
p 'fe
i-ED
fl>
^3
AiafG H)
3
a
n
3*
fT
3
CfQ fD
Fig-1 3 3
Relationships between (from left to right) the pachytene idiogram of rice, the molecular linkage map, and the classical map. Positions
of the kinetochores are indicated by Os on the idiogram, dark areas on the molecular linkage map, and C on classical map.
Relationships between the molecular and morphological markers, where known, are indicated by dashed lines. (Khush e t a h , 1996),
. J7Z91G
•RZ5SB
' ftn44
' 1^3 i
10-2
|cW4|
7^
to
5'
u
3
a
O
in
3cn
ir
w
o
w

304 Rice Breeding and Genetics: Research Priorities and Challenges
S'
K
a
Í3
5 Í I'S Í

R.J. Singh and G.S. Khush 305
C4
u
bO
IS

306 Rice Breeding and Genetics; Research Priorities and Challenges
Monosomic Alien Addition Lines
The wild relatives of cultigens are used as sources of abiotic and biotic
stresses for crop improvement. This is accomplished by producing a
complete series of monosomic alien addition lines (MAALs). Earlier
attempts for producing wide crosses in rice ended at the Fj stage.
Shastry and Ranga Rao (1961) produced sterile hybrids between O.
sativa and O, australiensis, Seed sterility was due to extremely low
meiotic chromosome pairing at metaphase I. They proposed that sterility
was due to timing imbalance between chromosomes of the parents.
Shastry et at, (1961) also produced an interspecific hybrid between O.
sativa and O. officinalis. Pachytene chromosome pairing revealed a loose
association, and 58.5% sporocytes at diakinesis showed 24 univalents.
Attempts were not made to produce amphiploids in order to restore the
seed fertility. Failure to produce a large number of wide crosses in the
genus Oryza is due to prefertilization incompatibility (Sitch and Romero
1990).
Shin and Katayama (1979) crossed a synthesized autotetraploid O.
sativa {In - 4x - 48; genome AAAA) with O. officinalis {In = = 24; CC).
The allotriploid {In = 3x = 36; AAC) was backcrossed to O. sativa. The
selfed progenies of plants with 2« = 26 and 27 produced plants with 2n =
25 chromosome (MAALs). They grouped 82 twenty-five chromosome
plants into 1 2 morphologically distinct types. Meiotic chromosome
pairing at metaphase t revealed a low frequency (0 - 1 ) of trivalent
configuration while 1211 + II configuration was most common.
Jena and Khush (1989) also produced MAALs from an interspecific
hybrid between O. sativa {In = 2x - 24; AA) and O. officinalis (2 m = 2x ~
24; CC). They rescued Fj ( 2 m = 2x = 24; AC) plants through in vitro
procedures. The crossability rate ranged from 1.0% to 2.3%. The method
of isolating MAALs with a single chromosome of O. offwinalis and
complete chromosome complement of O. sativa is shown in Fig. 13.4.
Table 13.5 shows chromosome segregation among the BC2 progeny.
Plants with 2x + 2, 3, 4, 5, and 6 were highly sterile and were greatly
modified morphologically. Twelve morphologically distinct types of
MAALs were isolated. These MAALs expressed striking resemblance to
the O. sativa primary trisomics. This suggests that the O. officinalis and
O. sativa chromosomes have similar gene content and have homologous
genomes (Khush and Singh, 1991). Chromosome association in MAALs
was primarily 1211 + 11. Three sporcocytes of MAAL 3 showed 1 III + IIII
configuration. Female transmission rates ranged from 6 .6 % to 26.8% and
male transmission rate was observed for MAALs 6 , 9,10, and 12. They
isolated diploid lines with brown planthopper resistance (BPH). Jena et
al. (1992) examined 52 BC2 p8 introgression lines using 174 RFLP markers

R J. Singh and G,S. Khush 307
0. sativa
(AA)
X 0, officinalis
(CC)
Embryo rescue
F i X O. sativa
(AC) (AA)
BCi X 0. sativa
(AAC) (AA)
i
BCj (AA + 1C, AA + 2C .... AA + 6C)
i
Monosomic alien addition lines
Fig. 13.4
Scheme used to isolate monosomic alien addition lines with a single
chromosome of Oryza officinalis and complete chromosome complement of O.
sativa. 0ena and Khush, 1989).
and identified 28 putative O. officinalis introgressed chromosome
segments. These segments were found in 11 of the 12 rice chromosomes.
Multara et al (1994) isolated,, by using the procedure, of Jena and
Khush (1989), 8(MAAL-1, A, -B, -7, -9, -10, -11, -12) of the possible 12
MAALs containing one chromosome of O, australiensis and a complete
chromosome complement of O. sativa. The MAALs resembled morphologically
the primary trisomics of O. sativa. The alien chromosome
associated with the O. sativa chromosome in a trivalent configuration in
all MAALs. The frequency of trivalent association at metaphase I ranged
from 7,5% (MAAL-4) to 18.5% (MAAL-12). The female transmission of
alien chromosome ranged from 4,2% (MAAL^l) to 37.2% (MAAL-12).
Male transmission of alien chromosome was recorded in MÁAL-5
(1 .2 %), -9(3.2%), and -12(2.4%). They screened 600 BC2 F4 progenies for
resistance to BPH and bacterial blight (BB). Four lines were resistant to
BPH and one line was resistant to Race 6 of BB.

308
Rice Breeding and Genetics: Research Priorities and Challenges
Table 13.5
Chromosome number in BC, progeny of O. sativa x O. offtcinalis (Jena
and Khush, 1989).
Chromosome no. No. of plants (%)
24 25 (26.6)
25 40 (42.5)
26 11 (11.7)
27 10 (10.6)
28 4 (4.3)
29 3 (3.2)
30 1 (1.1)
Total 94 (100.0)
References
Aggarwal, R.K., Brar, D.S. and Khush, G.S., 1997. Two new genomes in the Oryza complex
identified on the basis of molecular divergence analysis using total genomic DNA
, hybridization, Mol. Gen, Genet. 254:1-12.
Benneth M.D. and Leitch, I.J. 1997. Nuclear DNA amounts in angiosperms—583 new
estimates. Ann. Bot. 80:169-196.
Bouharmont, J., Olivier, M. and Dumonte de Chassart, M. 1985. Cytological observations in
some hybrids between the rice species Oryza sativa L. and O. glaberrima Steud, Euphytica
34: 75-81.
Burnham, C.R. 1962. Discussion in Cytogenetics. Burgess, Minneapolis, MN.
Chang, T.T, 1976. The origin, evolution, cultivation, dissemination, and diversification of
Asian and African rices. EiUphytica 25:425-441.
Chang, T.T. 1985. Crop history and genetic conservation: Rice—a case study, him State]. Res.
59; 425-455.
Chen, X., Temnykh, S., Xu, Y., Cho, Y.G. and McCouch, S.R. 1997. Development of a
microsatellite framework map providing genome-wide coverage in rice {Oryza sativa L.).
Theor. Appl. Genet. 95; 553-567.
Chung, M.C. and Wu, H.K. 1994. Cytological identification of the chromosomes involved in
rice translocation lines, Theor. Appl. Genet. 88:956-964.
Dolores, R.C., Chang, T.T. and Ramirez, D.A. 1979. The cytogenetics of Fj hybrids from Oryza
nivara Sharma et Shastry x O. sativa L. Cytologia, 44:527-540.
FAO. 1997. Year Book. Production 19%, 51:52-64..
Fukui, K., Ohmido, N. and Khush, G.S. 1994. Variability in rDNA loci in the genus Oryza
detected through fluorescence in situ hybridization. Theor. Appl. Genet. 87:893-899.
Harlan, J.R. 1965. The possible role of weed races in the evolution of cultivated plants.
Euphytica 14:173-176.
Hermsen, J.G. Th. 1970. Basic information for the use of primary trisomics in genetic and
breeding research. Euphytica 19:125-140.
Hu, C.H. 1964. Further studies on the chromosome morphology of Oryza sativa. In: Rice
Genetics and Cytogenetics. Elsevier, pp. 51-61.
Hu, C.H. 1968. Studies on the development of twelve types of trisomics in rice with reference
to genetic study and breeding programme. /. Agric. Assoc. China (N.S.) 63: 53-71 (in
Chinese with English summary).
Ishiki, K. 1991. Cytogenetical studies on African rice, Oryza glaberrima Steud, 3. Primary
trisomics produced by pollinating autotriploid with diploid. Euphytica 55; 7-13.

"Tí
RJ. Singh and G.S. Khush 309
Iwata, N. and Omura, T. 1975. Studies on the trisomics in rice plants (Oryza sativa L.), III.
Relation between trisomics and genetic linkage groups. }pn. /. Breed. 25:363-368.
Iwata, N. and Omura, T. 1976. Studies on the trisomics in rice plants (Oryza sativa L.), IV. On
the possibility of association of three linkage groups with one chromosome./pn. J. Genet.
51: 135-137.
Iwata, N. and Omura, T. 1984. Studies on the trisomics in rice plants (Oryza satmaL.), VI. An
accomplishment of a trisomic series in Japónica rice plants. Jpn. J. Genet. 59:199-204,
Iwata, N., Omura, T. and Nakagahra, M. 1970. Studies on the trisomics in rice plants {Oryza
satim L.), 1. Morphological classification of trisomics. Jpn. /. Breed. 20; 230-236.
Iwata, N., Satoh, H. and Omura, T, 1984. Relationship between the twelve chromosomes and
the linkage groups (Studies on the trisomics in rice plants {Oryza sativaL.), V, Jpn. J. Breed.
34: 314-31.
Izawa, T. and Shimamoto, K. 1996. Becoming a model plant: the importance of rice to plant
science. Trends Plant Sei. 1:95-99.
Jena, K.K. and Rhush, G.S. 1989. Monosomie alien addition lines of rice: production,
morphology, cytology, and breeding behaviour. Genome, 32:449-455.
Jena, K.K., Khush, G.S. and Kochert, G. 1992. RFLP analysis of rice {Oryza sativa L.)
introgression lines. Theor, Appl. Genet. 84: 608-616.
Kamisugi, Y., Nakayama, S., Nakijima, R., Ohtsubo, H., Ohtsubo, E. and Fukui, K. 1994.
Physical mapping of the 5S ribosomal RNA genes on rice chromosome 11. Mol. Gen.
Ge«ei. 245:133-138. .
Katayama, t . 1977. Cytogénetical studies on the genus Oryza, IX. The F^ hybrids from
synthesized amphiploid x species with the BBCC genomes. Jpn, /. Genet. 52:301-307.
Katayama, T. 1995. Cytogenetical studies on the genus Oryza, XIV. Intergeneric
hybridizations betwene tetraploid Oryza species and diploid teersia species. Jpn. J. Genet.
70: 47-55.
Katayama, T. 1997. Cytogenetical studies on the genus Oryza, IX. The F^ hybrids from
synthesized amphiploid x species with the BBCC genomes. Jpn. J. Genet. 52:301-307
Katayama, T. and Ogawa, T. 1974. Cytogenetical studies on the genus Oryza, VII.
Cytogenetical studies on Fj hybrids between diploid O. punctata and diploid species
having C genome. Jpn. J. Breed. 24:165-168.
Katyama, T. and Onizuka, W., 1978. Dihaploid-BC plants produced in F^ plants between
synthesized amphiploid (BBCC) and diploid (BB)-or tetraploid (BBCC)-Ofyzfl punctata
Kotschy ex Steud. Jpn. J. Genet. 53: 67-70.
Katyama, T., Shin, T.B. and Onizuka, W. 1977. Cytogenetical studies on the genus Oryza, X.
Cytogenetics of tetraploid Fj plant between amphiploid punctata-eichingeri and BBCC
genome species. J. Fflc. Agrie. Kyushw Unit». 22; 99-105.
Khush, G.S. 1973. Cytogenetics ofAnmploids. Acad. Press, NY London.
Khush, G.S. 1975. Rice In: Handbook of Genetics. R.C. King (ed.). Plenum Press, Vol. 2, pp. 31-
58.
Khush, G.S. 1997. Origin, dispersal, cultivation and variation of rice. Plant Mol. Bioh 35:25-
34.
Khush, G.S. and Singh, R.J. 1991. Chromosome architecture and aneuploidy in rice. In:
Chromosome Engineering in Plants: Genetics, Breeding, Evolution. Part A, P. K. Gupta, and T.
Tsuchiya (eds.). Developments in Plant Genetics and breeding, 2A. Elsevier Science B.V.,
pp. 577-598.
Khush, G.S., Singh, R.J. Sur, S.C. and Librojo, A.L. 1984. Primary trisomics of rice: Origin,
morphology, cytology, and use in linkage mapping. Genetics 107; 141-163.
Khush, G.S., Singh, K., Ishii, T., Parco, A., Huang, N., Brar, D.S. and Multani, D.S. 1996.
Centromere mapping and orientation of the cytological, classical, and molecular linkage
map of rice. In; Rice genetics. III. G.S. Khush (ed.). IRRI, Manila, Philippine pp. 57-75.

310 Rice Breeding and Genetics: Research Priorities and Challenges
Kurata, N. 1986. Chromosome analysis of mitosis and meiosis in ricé. In: Ríce Genetics. Island
Publ. House, Inc., Manila, Philippines, pp. 143-152.
Kurata, N., Omura, T. and Iwata, N. 1981. Studies on centromere, chromomere and nucleolus
in pachytene nuclei of rice, Oryza sativa, microsporocytes. Cytologia 46: 791-800.
Librojo, A.L. and Khush, G.S. 1986. Chromosomal location of some mutant genes through the
use of primary trisomics in rice. In: Rice Genetics. Island Publ. House, Inc., Manila,
Philippines, pp. 249-255.
McCouch, S.R., Kochert, G., Vu, Z.H., Wang, Z,Y., Khush, G.S., Coffman, W.R. and Tanksley,
S.D. 1988. Molecular mapping of rice chromosomes. Theor. Appl, Gmet. 76:815-829.
Multani, D.S., Jena, K.K., Brar, D.S., de los Reyes, B.G., Angeles, E.R. and Khush, G.S. 1994.
Development of monosomic alien addition lines and introgression of genes from Orym
australiensis Domin. to cultivated rice O, sativa L. Theor. Appl. Genet. 88:102-109.
Nagao, S. and Takahashi, M. 1963. Trial construction of twelve linkage groups in Japanese
rice (Genetical studies on rice plant, XXVIIJ. Tacul. Agrie. Hokkaido Univ., Sapporo. 53:72-
130.
Nayar, N.M. 1973. Origin and cytogenetics of rice. Adv. Genet. 17:153-292.
Nezu, M ,,, Katayama, T.C. and Kihara, H. 1960. Genetic study of the genus Otyza. I.
Crossability and chromosomal affinity among 17 species. Seiken Ztho, 11:1-11.
Ñishimura, Y. 1961. Studies on the reciprocal translocations in rice and barley. Bull. Natl.
Inst. Agrie. Sei. Ser. D 9:171-235 (In Japanese with English summary).
Nonomura, K,, Yoshimura, A. and Iwata, N. 1997. Cytogenetical gene mapping by reciprocal
translocation and tertiary trisomic analysis in rice {Oryza sativa L.). Genes Genet. Syst. 72:
41-49.
Ogawa, T. and Katayama, T. 1973. Cytogenetical studies on the genus Oryza, VI.
Chromosome pairing in the interspecific hybrids between O. officinalis and its related
diploid species. Jpn. J. Genet. 48:159-165.
Ogawa, T. and Katayama, T.1974. Cytogenetical studies on the genus Oryza. Chromosome
pairing in the interspecific hyrbid between diploid O. punctata and O. officinalis Jpn. J.
Genet. 49; 257-260.
Ohmido, N. 1995. Ricé chromosomes studies by fluorescence in situ hybridization with
special reference to physical mapping and chromosome structure./. F«c. Agrie. Holáaido
Univ. 66: 277-320.
Ohmido, N. and Fukui, K. 1995, Cytological studies of African cultivated rice, Oryza
glaberrima. Theor. Appl. Genet. 91:212-217.
Oka, H. 1964. Pattern of interspecific relationships and evolutionary dynamis in Oryza. In:
Rice Genetics and Cytogenetics. Elsevier, pp. 71-90.
Ranjhan, S., Glaszmann, J.C., Ramirez, D.A. and Khush, G.S. 1988. Chromosomal location of
four isozyme loci by trisomic analysis in rice {Oryza sativa L ). Theor. Appl. Genet. 75:541-
545.
Senchez, A.C. and Khush, G.S. 1994. Chromosomal location of some marker genes in rice
using the primary trisomics. /. Hered. 85: 297-300.
Sato, S. and Shinjyo, C. 1995. Identification of earliness genes derived from 15 reciprocal
translocation homozygotes of rice, Oryza sativa L. fpn.}. Breed. 45:45^9.
Second, G. 1982. Origin of the genic diversity of cultivated rice (Oryza spp.)'. study of the
polymorphism scored at 40 isozyme loci. Jpn.}. Genet. 57:25-57.
Seshu, D.V. and Venkataswamy, T. 1958. A monosomic in rice. Madras Agic, J. 45:311-314.
Shastry, S.V.S. and Ranga Rao, D.R. 1961. Timing imbalance in the meiosis of the Fj hybrid
Oryza sativa x O. australiensis. Genet. Res. Cambridge 2:373-383.
Shastry, S.V.S., Ranga Rao, D. and Misra, R.N. 1960. Pachytene analysis in Oryza, I.
Chromosome morphology In Oryza sativa. Indian J. Genet, and Plant Breed. 20:15-21.

RJ. Singh and G.S. Khush 311
Shastry, S.V.S., Sharma, S.D, and Ranga Rao, D.R. 1961. Pachytene analysis in Oryza^ III.
Meiosis in an inter-sectional hybrid, O. sativa xO. officinalis, The Nucleus, 4: 67-^0.
Shin, Y.B. and Katayama, T. 1979. Cytogenetical studies on the genus Oryza. XI. Alien
addition lines of O. sativa with single chromosomes of O. officinalis. }pn. J. Genet 54:1-10.
Singh, K., Multani, D.S. and Khush, G.S. 1996a. Secondary trisomics and telotrisomics of rice:
Origin, characterization, and use in determining orientation of the chromosome map.
Genetics 143:517-529.
Singh, K., Ishii, T., Parco, A., Huang, N., Brar, D.S. and Khush, G.S. 1996b. Centromere
mapping and orientation of the molecular linkage map of rice (Oryza sativaL.), Proc. Natl.
Acad. Sci. USA 93: 6163-6168.
Singh, R.J. 1993. Plant Cytogenetics, CRC Press; Inc., Boca Raton, PL.
Sitch, L. A. and Romero, G .0 .1990. Attempts to overcome prefertilization incompatibility in
interspecific and integeneric crosses involving Oryza sativa L. Genome 33: 321-327.
Vaughan, D.A. 1994. The Wild Relatives of Rice (A Genetic Resources Handbook). IRRI,
Manila, Philippines.
Wang, Z.X. and Iwata, N. 1991. Production of n + 1 plants and tetrasomics by means of
anther culture of trisomic plants in rice (Oryzii sativa L.). Theor, Appl. Genet. 83; 12-16.
Wang, Z.X. and Iwata, N. 1995. Aneuploids and tetrasomics in rice (OryzasativaL.) derived
from anther culture of trisomics. Genome 38:696-705.
Wang, Z.Y., Second, G. and Tanksley, S.D. 1992. Polymorphism and phylogenetic
relationships among species in the genus Oryza as determined by analysis of nuclear
RFLPs. Theor. Appl. Genet. 83:565-581.
Wang, Z.X., Iwata, N., Sukekiyo, Y. and Yoshimura, A. 1991. Induction of chromosome
aberrants in rice (Oryza sativaL.) by using irradiated pollen./. Fac. Agrie. Kyushu Univ. 36:
99-108.
Wang, Z.X., Ideta, O., Yoshimura, A. and Iwata, N. 1995a. Ideritification of extra
chromosomes of aneuhaploids and tetrasomics in rice and use of these aneuploids in
genome analysis. Breed. Sci. 45: 327-330.
Wang, Z.X., Kurata, N., Saji, S., Katayose, Y. and Minobe, Y. 1995b. A chromosome 5-specific
repetitive DMA sequence in rice (Oryza sativa L.). Theor. Appl. Genet 90:907-913.
Watanabe, Y., Ono, S., Mukai, Y, and Koga, Y. 1969. Genetic and cytogenetic studies on the
trisomic plants of rice, Oryza sativa. L., I. On the autotriploid plant and its progenies, Jpn.
J. Breed. 19:12-18.
Wu, K.S,, Glaszmann, J.C. and Khush, G.S. 1988. Chromosomal location of ten isozyme loci
in rice (Oryza sativa L.) through trisomic analysis, Biochem. Genet. 26:303-320.
Yoshimura, A., Iwata, N. and Omura, T, 1982. Linkage analysis by reciprocal translocation
method in rice plants (Oryza sativa L,), III. Marker genes located on chromosomses 2,3,4,
and 7. jpn }, Breed. 32: 323-332.
Yu, Z.H., McCouch, S.R., Kinoshita, T., Sato, S. and Tanksley, S.D. 1995. Association of
morphological and RFLP markers in rice (Oryza sativa L.). Genome 38:566-574.

14
Species of Genus Oryza and
Their Interrelationships
S.D. Sharma^, S.R. Dhua^ and P.K. Agrawal^
INTRODUCTION
The genus Orym belongs to the family Poaceae, tribe Oryzeae. The genus
is characterized by the presence of two rudimentary glumes in the
pedicel below the point of articulation, two sterile lemmas (generally
small) above the point of articulation (absent in perrieri, tissermti and
anqustifolia), a single tough fertile lemma having five vascular bundles
and ending in an awn (awn absent in O. meyeriana and in many cultivars
of saliva), a palea having three vascular bundles but similar in shape and
texture to that of fertile lemma, six stamens and a bifid feathery stigma
(Tateoka, 1962a). Based on these criteria, Oryza coarctata was excluded
from the genus Oryza (Tateoka, 1962a; Sharma and Shastry, 1966). Later
Launert (1965) excluded O. perrieri, O, Hsseranti and O. aiigustifolia from
the genus but, following Tateoka (1962a), we have included these three
species in the genus Oryza.
Monographs on the genus Oryza have been written by Prodoehl
(1992), Roschevicz (1931), Chevalier (1932) and Vaughan (1994). The
monograph written by Roschevicz (1931) is the most comprehensive and
that by Vaughan (1994) the most recent. The species of the genus Oryza
M.S. Swaminathan Research Foundation, Chennai 600113.
^ Central Rice Research Institute, Cuttack 753 006.

314 Rice Breeding and Genetics: Research Priorities and Challenges
have been enumerated by Tateoka (1963), Chang ((1985), and Vaughan
(1994).
TAXONOMY OF GENUS OKYZA
The genus Oryza consists of about 26 species. The exact number of
species is highly subjective and differs from author to author. The
species are distributed in the tropical and subtropical regions of the
world. They are either hydrophytes growing in open sunshine or
mesophytes adapted to moist soil and growing in partial shade. They
are either diploid (2n = 24) or tetraploid {In = 48), forming bivalents only
at meiotic metaphase-I. A list of Oryza species following Tateoka (1963),
Chang (1985) and Vaughan (1994) with slight modification (as discussed
later) is presented in the Appendix.
Classification
The species of Oryza were grouped into sections by Roschevicz (1931)
and Chevalier (1932). Later, Sharma and Shastry (1965c) classified the
genus into three sections and nine series, which were later reduced to
only eight by Sharma and Sampath (1985)' Recently, Vaughan (1994) has
grouped the species into species complexes, which more or less
correspond with our series. In this review, Sharma and Shastry's (1965c)
classification as given below has been followed. (The figures in
parentheses indicate the number of species in that series).
Section
Oryza
Angustifolia
Padia
Classification of genus Oryza
Series
Sativae (8)
Latifoliae (10)
Brachyanthae (2)
Perrierianae (2)
Meyerianae (4)
Ridleyanae (2)
Schlechterianae (1)
Representative Species
O. sativa
O. latifoHa
O. brachyantha
O. perrieri
O. meyeriana
O. ridleyi
O. schlechteri
SÈC T . O r y z a
The species of this section form the largest group in the genus and are
distributed in the tropics of the Old as well as the New World. This
section has been divided into two series, viz. Ser. Sativae and Ser.
Latifoliae.

S.D. Sliarma e t a l. 315
Ser. sativa
This group is represented by two cultivated species^, tiamely/ O. sativa of
South and Southeast Asia and O. glaberrima of western Africa and their
wild relatives. The Asian cultivated rice; O, sativa, is now widely
distributed and is cultivated in all the rice-growing areas of the world
including the homeland of O. glaberrima.
In Asia; the series is represented by a perennial wild species (O.
rufipogon), an annual wild species (O. nivara), and the cultivated species
(O. sativa). In addition, hybrids between the cultivated and the two wild
species also occur in nature and are commonly referred to as O, sativa f.
spontanea. A brief description of these taxa is given below.
O, rufipogon: a procumbent plant spreading on the ground and
branching at internodes. It grows up to 3 m long in swamps. When the
water level rises, the branches remain suspended in water. The panicle
emerges above the surface of water with a long exsertion. The
inflorescence is a panicle with lax primary branches, which bear a few
secondary branches. The spikeiets are long, slender, with flexuous awns.
The anthers are long and completely fill the spikeiets. The stigma is
generally purple in color and protrudes out of the lemma and palea after
anthesis. It is a photosensitive species flowering during November/
December. It is widely distributed from India and Sri Lanka to southern
China, Vietnam, and Indonesia. It grows in coastal lowlands,
particularly in deltaic areas of rivers of these countries.
O. nivara: an annual species with semierect habit, growing generally
to a height of less than a meter. It grows in seasonal pools or in the
margins of tanks up to a depth of 0.5 m of water. The panicle is poorly
exserted or even partly inserted in the sheath of the flag leaf. When
compared with O. rufipogon, the panicle has short primary branches and
bolder spikeiets. Awns are longer, bolder, and stiffer than its perennial
counterpart. It is a photoperiod insensitive species and flowers during
August to October. It is specially adapted to the plateau regions of India,
Myanmar, Thailand, and southern China.
In Australia, the annual wild species, O. meridionalis, is distributed
in the tropical areas. It is a tall, erect species growing in seasonal ditches
and pools of water.
O. sativa: The cultivated species originally grown in South and
Southeast Asia but now in cultivated in all continents of the world
(except Antarctica). It is an armual species but has potential to perennate.
The main feature that distinguishes this species from the wild rices is
that the spikeiets of the former do not shatter after maturity. The
additional characters are synchronous flowering of all the tillers, higher
number of spikeiets per panicle, variation in color (straw, brown.

316 Rice Breeding and Genetics; Research Priorities and Challenges
golden^ etc.) of the lemma and palea especially after maturity, often lack
of awns, white pericarp, etc.
O. sativa t spontanea: The wild rices that grow in cultivated fields
resemble cultivated species except that the spikelets shatter at maturity.
The additional characters that may or may not be present are black husk,
awn, red pericarp, etc. These spontanea rices are "wild" in the sense that
they shed their spikelets on maturity. Biosystematically, these are the
result of introgression of rufipogon or nivara genes into O. sativa. Eradication
of these spontanea rices poses a problem for the farmers because of
their close resemblance to cultivated rices, especially at the vegetative
stage.
The African continent provides more or less a parallel situation.
There are three species: one perennial, one annual, and one cultivated.
Besides, there are wild forms that resemble the African cultivated rice
O. glaberrima, often segregate on selfing and hence cannot be assigned to
any taxonornic status. The perennial wild rice of Africa {longistaminata)
rarely takes part in the natural hybridization with the African cultivated
rice {galberrima) and, in this sense, the African situation differs from the
Asian one. A brief morphological description of the three African species
is provided below.
O. longistaminata is a stoloniferous plant with erect habit. The ligule
is long, pointed, and bifid as in O. rufipogon. The spikelets are long and
slender and the anthers fill the spikelets very well. It is generally
assumed that the species has developed a self-incompatibility system. It
is widely distributed throughout tropical Africa including Madagascar
where it grows in perennial swamps,
O. harthii (syn. O. hreviligulata) is a plant of medium height and
bushy habits. The spikelets and awns of this species are the longest in the
genus. The armual wild species, O. barthii, is confined to the northern
part of tropical Africa and is distributed from Senegal to Sudan.
O. glaberrima is the cultivated species of rice in Africa and is
characterized by a glabrous leaf and lemma palea. It is distinguished
from O. sativa by a papery obtuse ligule and lack of secondary branching
of the panicle. This cultivated species (glaberrima) is grown from Sierra
Leone and Senegal to Cameroon. In recent years, the Asian cultivated rice
O. sativa has been replacing. O. glaberrima because of better agronomic
adaptability and yielding potential of the former.
O. glaberrima f. stapfii is the result of introgression of barthii characters
into O. glaberrima as the genetic barrier between the two species (barthii
and O. glaberrima) is incomplete and natural hybridization between the
two species does take place. These stapfii rices of Africa are similar to
O. glaberrima but they shatter their spikelets on maturity. Besides, they
have hairy spikelets, leaves, and awns as seen in O. barthii.

S.D. Sharaia et ah 317
The ligules of O, longistaminata are long, pointed, and bipartite as in
the species of Ser. Satime of Asia and America. This is an important
taxonomic character and, in this sense, the species is closer to the Asian
and American species and distinctly different from the other two {barthii,
glaberrima) African species. In other words, among the African elements,
O. longistaminata does not seem to be as closely related to O. glaberrima
and O. breviligulata since, in Asia, O. rufipogon is related to O' satim and
O. nivara. Besides, O, longistaminata rarely takes part in natural hybridization
with O. glaberrima and, in this sense, differs from the Asian
situation. Genomically. O. longistaminata represents a subgenome (A^A^)
that differs from that (A^A®) of O. barthii and O. glaberrima.
In America, a single species, O. glumaepetula, is widely distributed
from Cuba to Paraguay.
All the species of Ser. Sativae are diploid with 2n = 24 chromosomes
and the genomic constitution of all these species is the same (AA)
though subgenomic differentiation within the genome has been
recognized (Yeh and Henderson, 1961, 1962).
Nomenclatural confusion
The delimitation of species and hence the nomenclature of various taxa
of Ser. Sativae have been debated subjects. The three perennial elements
{rufipogon, longistaminata, and glumaepetula) were earlier treated as three
subspecies of a single species and were identified as O, perennis subsp.
halunga, O. perennis subsp. barthii and O. perennis subsp. cubensis
respectively (IRRI, 1964a). Some authors consider the American element
(glumaegetula) a mere variant of O. rufipogon (Second, 1982). The name
O. barthii was 'earlier mistakenly used for the perennial rice (O.
longistaminata) until Clayton (1968) clarified that it is the correct name
for the annual wild rice (known earlier by the synonym O. briviligulata).
The Australian annual wild rice, O. meridionalis, was earlier treated as a
variant of O. rufipogon (Morishima, 1984) or of O. nivara (Sharma and
Shastry, 1965a).
Genetic barriers
The genetic barrier among O. rufipogon, O. nivara and O. sativa is not
complete. Natural hybridization takes place particularly between O.
sativa and O. rufipogon in coastal areas and between O. sativa and O.
nivara in plateau regions of South and Southeast Asia. This leads to
introgressive hybridization and transfer of genes of one species into the
other. The occurrence of introgressed forms in nature has blurred the
distinction of these species leading to differences of opinion with regard
to their delimitation and nomenclature.

318 Rice Breeding and Genetics: Research Priorities and Challenges
Similarly, the genetic barrier between O. glaberrima and O. barthii is
incomplete and hence natural hybridization between these two species
in Africa has resulted in introgressive hybridization and transfer of
genes of one species into the other. This has resulted in many
intermediate forms and has obscured the distinction of these two
species.
O. saliva and O. glaberrima show many morphological similarities
and parallel variation. However, the Fj hybrids between these two
species are completely sterile O. meridionalis of Australia is genetically
completely isolated from the other species of this group and the Fj
hybrids of this species with other species of Ser. Sativae are completely
sterile.
The three pereimial species, O, rufipogon of Asia, O. longistaminata of
Africa, and O, glumaepetula of America, were earlier treated as a single
species and identified as O. perennis following Chevalier (1932) and
Chatterjee (1948). Of these, hybrids between O. longistaminata and the
other two species are completely sterile. On the other hand, the hybrid
between O. rufipogon and O. glumaepetula is partially sterile and hence
some taxonomists merge the latter with the former.
Ser. Latifoliae
Ser,Latifoliae comprises eleven species, five of which are diploid (2 n =
24) and six tetraploid (2 n = 48). This group is distributed throughout the
tropics and subtropic^ of the world and often adapted to partial shade of
forests. A better understanding of this group has come from the study of
their chromosome numbers and genomic constitutions. Despite the
variation in their ploidy and genomic constitution, the morphology of
species often overlaps, so much so that this whole group has been
designated by Tateoka (1962b) and Vaughan (1994) as the O. latifolia
complex. A brief description of the species of this group is provided
below.
Diploid species
All the five diploid species of Ser. Latifoliae are distributed in Africa,
Asia, and Australia only.
O. officinalis is a large-sized plant with well-ramified panicle. It is
very similar to O. latifolia (an American tetraploid species) and was
misidentified as O, latifolia until their ploidy and genome differences
were established. It is widely distributed from the west coast of India to
Indonesia. This species grows in partial shade in forests near water
streams or in moist grounds.
O. rhizomatis is known only from Sri Lanka. It was collected,
described and so named by Vaughan (1989). The only other species of

S.D. Sharaia et al. 319
Ser. Latifoliae available in Sri Lanka is O. eichingeri. Compared to O.
dchingeri of Sri Lanka, O. rhizomatis is taller vyith larger spikelets.
Compare to O. officinalis (which it closely resembles), the spikelets are
smaller (Vaughan, 1989).
O. eichingeri is a small-sized plant adapted to humid tropical forest
of Uganda and Sri Lanka. It is the only species of Ser. Latifoliae that is
distributed in two continents, Africa and Asia. The Asian (Sri Lankan)
form was available in living form from the early years of cytogenetic
studies. The African element was first collected from Uganda in living
form in 1964 by Tateoka.
O. punctata is another diploid species of Ser. Latifoliae available in
Africa. Its spikelet size is larger than that of O. eichingeri. The plant is
adapted to open habitat and grows in seasonal ditches in Sudan and
adjoining Ethiopia. It was collected in living form for the first time by
Tateoka (1964).
O. australiensis is a tall plant with rhizomatous habit and has a wellramified
panicle. The spikelets are the largest in Ser. Latifoliae. This
species is distributed in northern parts of Australia. It grows in open
habitats in pools of water.
Tetraploid species Asian and African (tetraploid) species: Three
tetraploid species of Ser. Latifoliae are available in Africa and Asia. Of
these, O. minuta is confined to the Philippines and O. malampuzhaensis to
southern India only. O. schweinfurthiana is widely distributed in tropical
Africa. The plant size, panicle morphology and spikelet size and shape
do not offer many distinctive features in the three species. All three
species are tetraploid and have the same genomic constitution (BBCC).
American tetraploid species: O. latifolia, O. alta, and O. grandiglumis
are the only three species of Ser. Latifoliae available in the American
continent. All the species are tetraploid (2 n = 48) and have the same
genomic constitution (CCDD). All the three species are tall, robust
plants with well-ramified panicles. The three species are easily
identifiable. O. latifolia has the shortest spikelet among the three. Of the
latter two species, O. grandiglumis has larger sterile lemmas covering
almost the whole of the fertile lemma and palea. However, the
morphological characters of the three species overlap and the genetic
barrier among the three species is not complete. This has led some
taxonomists to suggest that the three species should be merged into a
single species. The distribution of O. grandiglumis is confined to Brazil
and its adjacent areas, whereas O. latifolia spreads up to Mexico and
Cuba in the north and up to northern Argentina and Paraguay in the
south, O. alta has an intermediate distribution.

320 Rice Breeding and Genetics: Research Priorities and Challenges
it I
Nomenclatural confusion
O. officinalis {2n = 24) was identified as O. latifolia {In = 48) by Hooker
(1897) and as O. minuta {In = 48) by Bor (1960). 0 . eichingeri of Sri Lanka
was identified as "O. officinalis (Ceylon)" by Morinaga and Kuriyama
(1959). It was treated as a separate species by Sharma and Shastry (1965)
and named O. coltina. The correct identity of this Sri Lankan form as (O.
eichingeri) was provided by Tateoka (1962a, 1963). This has been
supported by Biswal and Sharma (1987) and Vaughan (1994). A variant
of O. alia was wrongly referred to as O. paraguainensis by Morinaga and
Kuriyama (1960). Morinaga (1964) and Li and his associates (Li, 1964, Li
et ah, 1961; Wuu et ah, 1963). O. schweinfurthiana was misidentified as O,
eichingeri by many cytogeneticists until Tateoka collected the correct
material (O, eichingeri) from Uganda and made it available to rice
researchers. O. schweinfurthiana {2n = 48) and O. punctata {2n = 24) have
been treated as two different species by Roschevicz (1931) and Andrews
(1956) but Tateoka (1962) merged the former with the latter and, to
■differentiate them, referred to them as O. punctata {2x) and O. punctata
{4x). O. rmlampuzhaensis (2n = 48) has been treated as a subspecies of O.
officinalis {2n = 24) by Tateoka (1962).
Variation in Ser. LatifoUae
Ser. LatifoUae provides^ much variation in size of plant, ligule characters,
size and shape of panicle, and spikelets. In Ser, LatifoUae, O. eichingeri
represents the smallest size of plant. Plant size is largest in O.
australiensis, O. officinalis, and the three American tetraploid species
{latifolia, alta, and grandiglumis).
O. eichingeri grows in hot, humid atmosphere under forest shade in
well-drained soils. O. officinalis grows in partial forest shade often near
rurming streams. The ecological preference of all tetraploid species is
more or less similar to that of O. officinalis. O. punctata and O.
australiensis are adapted to open habitats.
The ligules in all the species of Ser. LatifoUae are truncated. In the
American tetraploid species, the ligules are fringed and hence this
character works as an identifying character for these species.
In O. officinalis, O. australiensis, and the American tetraploid species,
the primary branches form a whorl at the base of the. panicle and the
basal regions of the lowermost primary branches are naked. They do not
bear a secondary branch.
The spikelets of O. rhizomatis are characterized by a wash of purple
pigmentation. This character works as an identifying character for this

S.D. Sharma et al. 32%
species. In O. officinalis, the base of the sterile lemma is often pigmented^
which helps in the identification of this species.
The size of the spikelets is the largest in O, australiensis followed by
that of O. punctata, O. grandiglumis, and O. alia. The smallest spikelet
size is seen in O. minuta and O, eichingeri.
Sect, Angustifoiia
This section consists of four species, namely, O. brachyaniha, O.
angustifoiia, O, perrieri, and O. tisseranti. The section was divided into
two series, namely, Ser. Brachyanthae and Ser. Pettievanae by Sharma and
Shastry (1965 c). All the species of this section belong to tropical Africa
and are small-sized plants. Except O. brachyaniha, all the species are
restricted to small localities only.
Ser. Brachyanthae: O. brachyaniha is a small plant growing erect up to
a height of 30-70 cm. The culm is slender and leaves are 15-30 cm long
and 4-6 cm wide. The inflorescence is a racerrie, the spikelets are long
and cylindrical, and awns up to 10 cm long, robust and scabrid, O.
brachyaniha is widely distributed in the whole of tropical Africa from
Senegal to Sudan in the north and Zambia in the south. It occurs in
shallow ditches and margins of ponds in open habitat. It is a diploid (2n
- 24) species and its genome (FF) is different from any of the species of
Sect. Oryz« (Li, 1964).
O. angustifoiia has not been available to rice researchers in living
form. According to Launert (1965), it occurs in Ubangi-Shari, Congo
(Katanga), Kenya and Zambia (Kasama Dist.). Compared with O.
brachyaniha, O. angustifoiia has wider leaf blades, and chartaceous and
flexuous awns. It lacks sterile lemmas.
Ser. Perrierianae is represented by two species, O. perrieri and O.
tisseranti. Both species are perennial, slender, awned and lack sterile
lemmas. O. perrieri is geniculate and about 30 cm in height. It is occurs in
the vicinity of Majunga lake in Madagascar. O. tisseranti is comparatively
taller, has a rigid culm, larger leaves and more ramified panicles. The
spikelets are comparatively more cylindrical and slightly longer. Both
species are diploid with 2n = 24 chromosomes (Chang, 1985).
Laurnet in 1965 rerrioved O. angustifoiia, O. perrieri, and O. tisseranti
from the genus Oryza and placed them in the genus Leersia mainly
because they lacked sterile lemmas at the base of fertile lemma and
palea. This factor was duly considered by the original authors (Camus,
1926; Chevalier, 1932; Hubbard, 1950) while placing them in the genus
Oryza.

322 Rice Breeding and Genetics: Research Priorities and Challenges
Sect. Padia
This is a Southeast Asian group of species represented by perennial
plants often growing in shady habitats. This section comprises three
series, namely, Ser. Meyerianae, Ser. Kidleyanae, and Ser. Schlechterianae.
Ser. Meyerianae is represented by four species (or subspecies of a
single species?), O. abromeitiana, O. meyeriana, and O. granulata, which
are very closely related. Tateoka (1962b) recognized a single species O.
meyeriana and assigned subspecific ranks to the three elements. They are
all small plants (< 1 m), prefer shady and moist habitat, and are
distributed in South and Southeast Asia. All of them are diploid (2n -
24). Recently, Ellis (1985) reported one more species, O. indandamanica
from the Andaman Islands. It is suspected that O. indandaminica is
merely a variant of O. meyeriana sensu lato as recognized by Tateoka
(1963).
Ser. Ridleyanae consists of two species, namely, O. ridleyi and O.
longiglumis. Compared to O, meyeriana, the two species are taller and
grow in partial shade in Southeast Asia. The surface of their fertile
lemma is smooth, the inflorescence is a panicle, and the fertile lemma
ends in an awn. The difference between these two species is mostly in
quantitative characters only. The latter expresses reduction is spikelet
size, elongation of awns, and more setaceous nature of sterile lemmas.
While the former is widely distributed in Southeast Asia from
Tennasserim (Myanmar) to New Guinea, the latter has been reported
from New Guinea only.
Ser. Schlechterianae is represented by a single species, O. schlechteri. It
has been described in detail by Pilger (1914) and Roschevicz (1931).
According to their description, O. schlechteri is a small plant, about 40
cm tall. The spikelets are small (1.75 mm), awnless and have a smooth
surface of fertile lemma and palea. Recently, Vaughan et al. (1991)
collected this in living form from the blanks of the river Yamu in Papua
New Guinea. It is a creeping, stoloniferous plant. It grows in partial or
complete shade in mountains with primary forest cover.
CYTOGENETIC APPROACHES TO SPECIES RELATIONSHIPS
The species of Oryza are either diploid with 2n = 24 chromosomes
forming 12 II or tetraploid with 2 « = 48 chromosomes forming 24 II at
meiotic metaphase-I. The tetraploid species appear to be amphidiploids
only as they form 24 II with hardly any univalents, trivalents or
quadrivalents. The chromosome numbers of all the Oryza species are
provided in Table 14.1.

S.D. Sharma et ah 323
Table 14.1
Species of Oryza, their chromosome numbers and genomes.
Species Chromosome number (2n) Genome
0 . alta 48 CCDD
0 . angustifolia 24 not known
0.au stralien sis 24 EE
0 . barthii 24 AA
0 , braehyantha 24 FF
O .eichingeri 24 CC
O. glaberrim a 24 AA
O. glum aepetula 24 AA
O. grandiglum is 48 CCDD
O. granúlala 24 not known ;i
O. M ifolia 48 CCDD 1
O. iongiglum is 48 not known
0 , longistam inata 24 • AA !
0 . malampuzhaensis 48 BBCC
0 . m eridionalis 24 AA
O. m eyeriana 24 not known ■
O. minuta 48 BBCC
0 . nivara 24 AA
0 . officinalis 24 CC(orDD*)
O, perrieri 24 not known
O. punctata 24 BB
O. ridleyi 48 not known |
O. rufipogon 24 AA ;
0 . sativa 24 AA !
0 . schlechteri 24 not known
0 . schw einfurthiana 48 BBCC
0 . tisseranti 24 not known
* Known to be CC but proposed in this paper to be DD.
f
From 1938 to 1943; Morinaga reported pairing behavior of chromosomes
of O. sativa, O. minuta, and 0 . M ifolia as well as of their
hybrids. The three parental* species are fertile (F) but their interspecific
hybrids were completely sterile (S). Based on the pairing behavior of
chromosomes in thé Fi^ hybrids, he inferred their genomic constitution
as follows:
0 . saiiva (F) 12 II AA
0 , minuta (F) 24 II BBCC
0 . latifoUa (F) 2411 CCDD
sativalminuta (S) 361 ABC
sativaAatifolia (S) 361 ACD
minuta/lg.tifoUa (S) 12 II + 24 I BCCD
This proposition presumed the occurrence of diploid species with
BBy CC; and DD genomes in nature followed by their interspecific
hybridization and amphidiploidy in the evolution of tetraploid species.
In subsequent studies on genome analysis, these three species {sativa,

324 Rice Breeding and Genetics: Research Priorities and Challenges
minuta, latifoHa) virtually became the "tester" species in the sense that
the genomes of other species were tested against the genomic
background of these three species.
AA Species
Morinaga and Kuriyama (1956^ 1957, 1960) reported the hybrids of O.
sativa with O. glumaepetula (their "O. cuhensis'% O. glaberrima, O. harthii
(their "O. breviligulata'') and O. rufipogon (their "0 . perennis"). The first
three hybrids were completely sterile and the last one partly fertile.
Normal pairing of chromosomes with 12 II was observed in each case.
They concluded that all these species have the same (AA) genome. Nezu
et at. (1960) crossed O. sativa with stapfii and spontanea types. While the
former hybrid was completely sterile, the latter was fully fertile.
However, the chromosomes in the hybrids formed 1 2 II during meiosis.
This proved that these two taxa also had the same genome (AA) as O.
sativa. Yeh and Henderson (1961) crossed O. sativa with O. longistaminata
(their "O. barthii"). The hybrid was sterile although 12 II were formed in
meiosis. This established the genome of O. longistaminata as AA.
The findings of all these authors conclusively established that the
genome of all the species of Ser. Sativae is AA. Supporting evidence
came from the interspecific hybrids involving AA species and other
BBCC and CCDD species.
BBCC Species
!t
Morinaga and Kuriyama (1960) reported the hybrid schweinfurthiana (24
n ) X minuta (24 II) as forming 24 II. It was therefore inferred that O.
schweinfurthiana (their "O. eichingeri") has the same genome (BBCC) as
O, minuta. This was subsequently confirmed by Hu (1970).
Gopalkarishnan, et al. (1965) reported obtaining hybrids of O.
malampuzhaensis with O. minuta and O. schweinfurthiana. The hybrids
were sterile and formed 24 II in majority of the PMCs. It was inferred
that O, malampuzhaensis has the same BBCC genome as O. minuta and O.
schweinfurthiana. This established that all the African and Asian
tetraploid species of Ser. Latifoliae have the BBCC genome.
CCDD Species
Morinaga and Kuriyama (1960) crossed O. latifolia with O. alta (their "O.
paraguaiensis") and observed 2 4 II in the F;i hybrid. They concluded that
the latter also has the CCDD genome as in the former. Li et al. (1962)

n
S.D. Sharma et al. 325
repeated this cross and confirmed this view. Nezu et al. (1960) crossed
O. latifolia with O. alia and observed 24 II in the hybrid. This again
established that the genome of O. alia is CCDD. Morinaga (1964)
reported the hybrids of O. grandiglumis with O. latifolia and O. alia (his
"O. paraguaiensis") and showed that G. grandiglumis also has the same
genome (CCDD). These studies established that all the American
tetraploid species of Ser. Latifoliae {latifolia, alta, grandiglumis) have the
CCDD genome.
Confirmatory Crosses
Crosses between AA species and BBCC species have been made by
Sharma and Seetharaman (1955), Morinaga and Kuriyama (1960), Nezu
et al. (1960), Kihara et al. (1961), Li et al. (1962), Bouharmont (1962), and
Hu (1970). In most of these crosses, 3 6 1 have been observed, confirming
the ABC nature of the hybrids to be that of sativa x minuta (Morinaga,
1940).
Interspecific hybrids involving AA species and CCDD species have
been reported by Gotoh and Okura (1935), Hirayoshi (1937), Morin and
Kuriyama (1960), Moringa, et al. (1960,1962), Nezu et al. (1960). Kihara et
al. (1961) and Li et al. (1962). In most of these crosses, 3 6 1 were observed,
confirming the ACD nature of the hybrids to be that of sativa x latifolia
(Morinaga, 1941).
Crosses between the African tetraploid (genome BBCC) and the
American tetraploid (genome CCDD) species have been reported by
Morinaga (1964), Nezu et al. (1960), Kihara et al., (1961) and Li et al.
(1961, 1962). In all these hybrids 12 II + 24 I have been reported as
expected in BCCD hybrids typical of minuta x latifolia of Morinaga
(1943).
Genome of O. eichingeri (Sri Lankan form)
O. eichingeri occurs in equatorial Africa was well as in Sri Lanka, The
former was identified as "O. officinalis (Ceylon)" by Morinaga (1959),
Gopalkrishnan (1966) and Katayama et al. (1972). Sharma and Shastry
(1965) recognized it as a separate species and called it O. collina. Tate oka
(1962) identified it as O. eichingeri. Sampath and Subramanyam (1968)
crossed O. eichingeri of equatorial Africa with O, collina and reported the
hybrid to be partly fertile. They observed regular pairing of
chromosomes forming 12 11. It was interpreted that the genomes of both
the taxa are the same. The plant morphology, ecology, and genomic

i
j
326 Rice Breeding and Genetics: Research Priorities and Challenges
constitution support Tateoka^s (1962) view that the Sri Lankan material
(O. collina) may be treated as amere variant of O. eichingeri only. Biswal
and Sharma (1987) agreed with Tateoka (1962) and likewise considered
it only a variant of O. eichingeri.
Morinaga and Kuriyama (1959) crossed the Sri Lankan form of O.
eichingeri (their "O. officinalis Ceylon") with AA, BBCQ and CCDD
species. The pairing behavior of chromosomes observed in the hybrids
is shown below:
sativa (AA)/"officinalis (Ceylon)" , 241
minuta {BBCC)/"officinalis (Ceylon)"
12 II + 12 I
"officinalis (Ceylon)"/latifolia (CCDD)
1211 + 12 I
"officinalis (Ceylon)"/grandiglumis (CCDD) 1 2 II + 1 2 1
The pairing behavior of chromosomes indicated that the genome of
O. eichingeri (their "O. officinalis (Ceylon)") differed from O. sativa (AA)
and was homologous to one of the genomes of O. minuta (BBCC) and O.
latifolia (CCDD). Hence Morinaga and Kuriyama (1960) proposed that
the genome of the Sri Lankan form of O, eichingeri, i.e., their "O. officinalis
(Ceylon)"/ is CC.
Genome of O. eichingeri (African form)
iiH
hîil il*
The African form of O. eichingeri was crossed with O. sativa (AA)/ O.
minuta (BBCC), and O. latifolia (CCDD) by Hu (1970). In sativa x
eichingeri, he observed 2 4 1 in 65% PMCs, 1 II + 2 2 1 in 22% PMCs, and 2 -
4 II in the remaining 13%. In minuta x eichingeri, 10 II + 16 I were
observed in a few pollen mother cells but in the others, restitution nuclei
were formed. In latifolia x O. eichingeri, the number of bivalents per cell
was 9 but more than 50% cells had 10 to 12 II. According to Hu (1970),
this indicated that the genome of O. eichingeri is common with one of the
genomes of O. minuta (BBCC) as well as O. latifolia (CCDD). In other
words, the genome of O. eichingeri appears to be CC.
O. punctata
O. punctata is a diploid species with 2n = 24 chromosomes forming 12 II
at meiotic metaphase-I. Katayama (1967) crossed this species with O.
minuta and O. schweinfurthiana (his "O, punctata~4x") and observed 1 2 II
+ 12 I at the meiotic metaphase-I of their hybrids in either case. He
therefore considered that the genome of O. punctata is available in O.
minuta and O. schweinfurthiana (his "O. punctata 4x") (both BBCC). Hence
he proposed the genomic constitution of O. punctata (2x) as BB. Hu
(1970) observed a mean configuration of 12 II in latifolia x O. punctata

S.D. Sharma et al. 327
hybrid and concluded that the genome of O. punctata is neither CC nor
DD and thus confirmed Katayama^s (1967) assumption that O. punctata
is BB.
DD Species
O. rhizomatis is a newly described diploid species endemic to Sri Lanka,
To determine its genome, Dhua (1994) crossed this species with O.
latifolia (CCDD), O. minuta (BBCC), and O. eichingeri (CC) and studied
the pairing of their chromosomes in the meiotic metaphase of their Fj
hybrids. He observed 10 or more trivalents in 60% PMCs in the hybrid of
latifolia X rhizomatis the mean configuration being 9.68 trivalents, 2.14 II
bivalents and 2.68 univalents. He concluded the genome of O, rhizomatis
to be either CC or DD, confirming the earlier views that the C and D
genomes were homologous (Richhaira, 1960) and capable of pairing and
forming up to 12 bivalents (Li et al., 1963; Shastry, 1965). In the hybrid
minuta x O. rhizomatis, he observed 22 or more univalents in 84% PMCs,
the mean configuration for all the PMCs being 0.1 III + 4.90 II + 25.71. He
interpreted the genome of O. rhizomatis to be DD and the pairing (mostly
bivalents) observed in this hybrid due to homology between the C
genome of O. minuta and D genome cf O, rhizomatis. In eichingeri x
rhizomatis, he observed 5 to 10 bivalents in 70% PMCs, the mean pairing
for all the PMCs being 5.2 II + 13.61 I. According to him, the genomic
constitution of both these species are different. In other words, the
genomic constitution of this hybrid is CD and pairing is due to homology
between C and D genomes.
Genome of O. Officinalis
The hybrid between O. sativa and O. officinalis has been studied by many
workers (Ramanujan, 1937; Nandi, 1938; Gopalakrishnan, 1959;
Morinaga and Kuriyama, 1960; Shastry et al., 1961). They reported 24 I,
indicating that the genome of O. officinalis differs from that of O. sativa
(AA). O. officinalis has been crossed with O. latifolia, O. alta, and O.
grandiglumis (all CCDD) by many researchers (Morinaga and Kuriyama,
1960; Nezu et al, 1960; Li et al, 1962). The hybrids were sterile and
formed mostly 12 II + 12 I during meiosis. This indicated that the
genome of O. offtcinalis is either CC or DD.
In the hybrid malampuzhaensis x officinalis, Gopalakrishnan (1959)
observed 2 to 36 univalents in 42 PMCs with the following frequencies:
Univalents: 2-8 10-12 13-17 18-22 23-36
Frequencies: 2 11 8 14 7
It is evident that 29 of the 42 PMCs had more than 12 I, while 21 of
them had 18 I or more. The mean number of bivalents observed by him
'I

328 Rice Breeding and Genetics; Research Priorities and Challenges
was 9 II. Using the genomic symbols 0^0^ for BB, 0^0^ for CC, and
0^0^ for DD, he wrote "the genome of both the species may be
represented as
(malampuzhaensis) and 0^0^ {pfficinalisY. He
concluded that the genome of O. officinalis was DD. Katayama (1967)
studied altogether 36 PMCs of minuta x officinalis and recorded 10 to 13
II (mean 11.82 II). According to Dhua (1994)^ this was due to the
homology between the CC genome of O. minuta and DD genome of O.
officinalis. The hybrids officinalis x eichingeri (CC) and officinalis x collina
(CC) form 12 bivalents at meiotic metaphase-I (Hu/ 1970). However^
mostly bivalents were observed in the amphidiploid of this hybrid
(Gopalakrishnan, 1964; Hu, 1970) leading one to suspect that the genome
of O. officinalis is different from that of O. collina (CC). The hybrid
between coUina-officinalis amphidiploid and O. latifolia was studied by
Sharma et al. (1974). It had 85% pollen fertility and formed 20.85 II at
meiotic metaphase-I. This led them to suspect that the genome of O,
officinalis could be DD.
Dhua (1994) crossed O. officinalis with O. rhizomatis and studied the
pairing behavior of their chromosome sin their hybrid. He observed
10 to 12 bivalents in 74% PMCs and concluded that the genome of O.
officinalis is the same as that of O. rhizomatis (DD).
Historically speaking, the genome of O. officinalis has been assumed
to be CC ever since Morinaga and Kuriyama (1959) proposed the
genome CC for O. officinalis (Ceylon), which is now identified as O.
eichingeri only. In fatt, as far back as 1959, Gopalakrishnan (1959)
proposed the genome DD for O. officinalis. Using the genomic symbols
0^0^ for BB, 0^0^ for CC and 0^0^ for DD and commenting about the
hybrid malampuzhaensis x officinalis, he wrote "thus the genome of both
these species may be represented as
{malampuzhaensis) and
0^0^ {officinalis)". In the Fj hybrid (O^O^O^), either autosyndesis
between 0 ^0 ^or allosyndesis between and or can take place.
According to Gopalakrishnan (1964), there is no need to search for
plants with the D genome outside the species of O. officinalis as one or
the other geograplrical race of O, officinalis might contain it. Similarly,
Sharma (1964) commented that "the DD species is not expected to differ
much from O. officinalis".
Genome of O. australiensis
From 1960 to 1963, the interspecific hybrids of O. australiensis with O.
sativa (AA), O. minuta (BBCC) and O. alta (CCDD) were reported by
Morinaga and his associates (Morinaga and Kuriyama, 1960; Morinaga
et al., 1960; Morinaga et al., 1962) and many others (Nezu et al, 1960;
Kihara et al, 1961; Li et al, 1961, 1963).

All these hybrids were completely sterile and the number of
bivalents observed were few and could be interpreted as autosyndetic
only. It was therefore concluded that the genome of O. australiensis is not
homologous to either A, B, C, or D and hence a new genome symbol EE
was proposed for this species (IRRb 1964),
Genome of O. bmchyantha
S.D. Sharma el al. 329
O. brachyantha was crossed with O, sativa (AA), O. minuta (BBCC), O,
alia (Labelled as "O. paraguaiensis") (CCDD) and O. australiensis (EE).
Success in these crosses was achieved after a very high number of
pollinations and embryo culture of the hybrid seeds. All these hybrids
were sterile. It is remarkable that the small plant size of O. brachyantha
was dominant over large plant size of other species. Univalents were
observed in all the hybrids indicating that the genome of O. brachyantha
was not homologous to A, B, C or Taking into consideration the wide
morphological differences with O. australiensis, it was presumed that
this genome could not be the same as that of O. australiensis. The genome
symbol FF was therefore proposed for this species (IRRI, 1964).
Thus, the genomes of all the species belonging to Sect. Oryza have
now been determined. In Sect. O. angustifolia, the only genome of O.
brachyantha has been determined to date. Genome analysis of species
included in Sect. Padia has not yet been attempted. Genome analysis of
species of the genus Oryza remains incomplete.
Integenomic Relationships
Interspecific hybrids with a genomic constitution of ABC, BCE, and BCF
help in analyzing the relationship of B and C genomes. Similarly,
interspecific hybrids with ACD, CDF, and CDF genomes help in
determining the relationship of C and D genomes. From the data on
chromosomal pairing reported on such hybrids by many authors (Nezu
et al, 1960; Kihara et al, 1961; Li et al, 1961,1962,1963, 1964; Katayama,
1966a, 1966b; Nowick, 1986), it appears that the genomes B, C, and D are
and, among the three genomes, C and homologous D show greater
homology than B and C (Shastry, 1965).
Autopolyploids have been induced in some species and interspecific
hybrids resulted in plants with a genomic constitution CCCC,
CCCCDDDD, BBBBCCCC, and CCCCBBDD. In all such cases, the
number of quadrivalents observed at meiotic metaphase-I is far less
than expected, resulting mostly in bivalents (Watanabe and Ono, 1965,
1966; Hu, 1967). According to Watanabe and Ono (1966), the D genome

330 Rice Breeding and Genetics: Research Priorities and Challenges
at higher poloidy levels is capable of suppressing not only intergenomic,
but also intragenomic pairings in particular in the formation of
multivalents.
SuBGENOMic D if f e r e n t ia t io n
Based on the intergenomic pairing observed in ABC^ ACD, and BCCD
hybrids, Gopalakrishnan (1959), Richhaira (1960), Gopalakrishnan and
Sampath (1966) treated the B, C, and D genomes as subgenomes and
designated them as O^, and respectively.
The isogenomic species exhibit differences in their morphology,
geographic distribution and ecological adaptation. The fact that
spéciation has advanced further at the isogenomic level is a matter of
evolutionary significance in the genus. The interspecific hybrids within
these isogenomic species have been studied by many workers. These
hybrids are often partially sterile, indicating that genetic differentiation
has taken place even within the same genomes.
Yeh and Henderson (1961, 1962) crossed the species of AA genome
in various combinations. Based on the pollen sterility of their hybrids,
the degree of pairing of their chromosomes at meiotic metaphase-I, and
irregularities in synapsis and disjunction of chromosomes, they assigned
different subgenomes to the AA species as follows:
AA p . sativa, O. nivara, O. rufipogon
A W O, glaberrima, O. barthii
A^A^ O. longistaminata
0 . glumaepetula
O. meridionalis
(The superscripts have been changed with the change in the names of
species. The subgenome for O. meridionalis has been added. For
subgenomic symbols one may refer to Change (1985) Vaughan (1989).
According to Gopalakrishnan and Sampath (1966), O, latifolia, O.
alia, and O. grandiglumis may be treated as three subspecies of a single
species, O. latifolia only.
Gopalakrishnan (1965) crossed different ecotypes of O. officinalis
among themselves and observed complete sterility in their F^ hybrids.
He, however, obseirved formation of 1 2 II at meiotic metaphase-I in these
Fj hybrids. He concluded that sufficient genetic differentiation has taken
placé at the chromosomal level in this species through this genetic
différentiation has not been associated with commensurate
morphological differentiation.

S.D. Sharma et aL 331
ISOZYME AND MOLECULAR STUDIES
In the present study^ the taxonomy and phylogenetic relationships of
species in the genus are discussed based on recent findings in these
areas:
1. Electrophoretic pattern of alcohol dehydrogenase
2. Fluorescence in situ hybridization
3. Restriction analysis of chloroplast DNAs
4. Restriction analysis of nuclear DNA
5. Repetitive sequences for genome specificity.
Electrophoretic Pattern of Alcohol Dehydrogenase
Alcohol dehydrogenase (ADH), an enzyme of the glycolytic pathway, is
highly conserved and therefore useful for studying species level
relationships. Variability of ADH can be studied both under aerobic and
anaerobic conditions. Richard et a/. (1986) reported increased expression
of ADH under aerobic conditions but observed no novel bands under
such conditions. However, the variation in the genus Oiyza with respect
to ADH expression both under aerobic and anaerobic condition is not
known.
Using ADH isozyme patterns. Second (1982) studied the variability
within a large number of accessions of O. sativa (japónica, indica and
javanica), O. glaberrima, and species of the O. perennis complex of Asia
and Africa. He studied the electrophoretic patterns of 13 different
enzymes in 1,948 strains and found that the ADH locus was
monomorphic.
Isozyrhe patterns of ADH can be studied by electrophoresis in
starch gels, by polyacrylamide gel electrophoresis (PAGE) and by
isoelectric focusing (lEF) in polyacrylamide gels. Grover and Pental
(1992) extracted ADH from germinating seeds and studied the ADH
profiles of 141 different accessions of 19 Oryza species by PAGE. Based
on ADH isozyme patterns, the genus Oryza was classified into six
different groups represented by:
(i) O. sativa, O. nivara, O, rufipogon (Asian and some American
types), O. galberrima, O. barthii, and O. longistaminata
(ii) O. officinalis, O. punctata, O. minuta, O. latifolia, O. alia, and O.
grandiglumis
(iii) O. australiensis
(iv) O. ridleyi and O. longiglumis
(v) O. hrachyantha, and
(vi) O, granúlala and O. meyeriana.
However, if another conservative marker (Rubisco-LS protein) is
taken into consideration, then the three species of the American

332 Rice Breeding and Genetics: Research Priorities and Challenges
continent, O. laiifolia, O. alia, and O. grandiglumis form one closely
related group that is distinct from other species of the O. officinalis
group. The position of some of the O. rufipogori accessions (American
types) and O. eichingeri vis-a-vis O. punctata was not clear. However, O.
punctata (4x), O. minuta, O. alia and O, grandiglumis and an ADH pattern
similar to that of O. officinalis—a species presumed to have contributed
one of the genomes of these species.
Isozymes are, however, tissue specific and affected by the environment
as well as the stage of development. Their limited number
prevents them from providing complete genome coverage (Smith and
Smith, 1990). Nevertheless, the use of isozymes remains a quick, cheap,
and easy method for a preliminary survey based on a few markers.
Variability detected through Fluorescence in situ Hybridization
TISH'
Rice chromosomes with a large 17s-5.8s-25s ribosomal RNA gene
(rDNA) array have been identified as satellite chromosomes by their
characteristics. The chromosomes are also recognized as nuclear
organizing regions (NOR chromosomes). Although the rDNAcontaining
chromosomes show conspicuous characteristics as satellite
chromosomes, they are sometimes difficult to identify morphological
when the copy number of the rDNA units at the locus is small. The in
situ hybridization method offers a way out of this impasse since it is
based on the detection of rDNA loci directly by molecular hybridization.
Although in situ hybridization is now widely employed in cytogenetic
analysis, it is time consuming and need strict experimental protocols. In
the PISH technique in conjunction with the imaging method, use of
thermal cycle and various post treatments are quite reproducible and
convenient. Fukui et al. (1994) took nine rice species, trisomics for
chromosome 9 and 10 of O. sativa and their original variety IR24, to
study the rDNA in Oryza by use of FISH. The genomes under this study
were A, A^^, B, C, E, F and CD. One or two rDNA loci were identified in
with species an A genome. The B genome species showed three rDNA
loci. Two C genome species showed either two or three rDNA loci.
Species with the E and F genomes exhibited either two or one rDNA loci
respectively. The number of rDNA loci in the D genome may be either
two or three as in the case of the C genome species.
The study by Fukui et al (1994) revealed variability in the number of
rDNA loci in eight diploid and one tetraploid species within the genus
Oryza. O. rufipogon, a putative ancestor of cultivated rice, has rDNA
variability which is similar to that of cultivated rice that has either one or
two rDNA loci. Varieties of temperate regions have one rDNA locus

S.D. Sharma ef al. 333
while those of tropical and subtropical regions have two rDNA loci.
favanica is sometimes referred to as tropical japónica. Two javanica
varieties showed two pairs of rDNA loci, indicating their similarity with
indica (Table 14.2). This result may be explained by the environmental
similarity of the areas where both javanica and indica varieties are grown.
Restriction Endonuclease Analysis of Chloroplast DNAs
Restriction endonuclease analysis of the chloroplast DNAs (ctDNAs)
has been used to determine the phylogenetic relationship between
closely related species or genera. Results of such analysis have revealed
that ctDNAs from related taxa exhibit wide variation even within the
same genus (Ogihara and Tsunewaki, 1988; Kishima et al. 1987). Ishii et
al, (1986) isolated ctDNAs from rice species of AA genome and
established the relationship among them with reference to differences in
lengths of restriction fragments of ctDNAs. Ichikawa et al. (1986) isolated
total DNAs from the same Orym species and detected restriction
patterns by Southern hybridization with ctDNA from O, sativa as probes.
T a b le 14.2
N um ber of rDN A loci detected in cultivated rice {Oryza sativa)
Species Genome Varietal group Variety nam e N um ber of
rD N A loci
O. sativa AA japónica N ipponbare 1
Aikoku 1
Tushim aakam ai 1
Tarizaohsien 1
Kouketsum ochi 2
C h 7 8 1
C h 7 9 1
indica Chinsurah Boro II 2
Kasalath 2
IR 3 6 2
javanica Ketan N anga 2
Inakupa 2
0. rufipogon AA Annual type 2
Perennial type 2
Perennial type 1
O. glumaepatula A8f AfiP 1
O. punctata BB 3
0. officinalis CC 3
O. eichingeri cc 2
O, australiensis EE 2
O. hrachyantha FF 1
0 , latifolia CCDD 5

334 Rice Breeding and Genetics: Research Priorities and Challenges
However, they used nine strains that represented only seven species of
Oryza. Similarly, Dally and Second (1990) isolated ctDNAs from 247
accessions representing 13 Oryza species and detected differences while
examining restriction patterns by agarose gel electrophoresis.
Chloroplast DNA from O. sativa was cloned by Hirai et al. (1985) and
a physical map was constructed. Subsequently, the complete sequence
of the ctDNA was determined by Hiratsuka et al. (1989). Kanno and
Hirai (1992) isolated ctDNA from O. punctata, O. officinalis, and O.
australiensis and constructed overlapping clone banks of the entire
chloroplast genome of each of these three species. They also constructed
physical maps of the ctDNA of these species and compared them. It was
found that two fragments of ctDNAs from O. punctata and O. officinalis
were shorter than the corresponding fragments from O. sativa.
Nevertheless, it is difficult to explain how these deletions/insertions
could be produced during the evolution of the various species of Oryza.
Qne may assume that mutational events had occurred minimal times
and deletions had been hardly reversible. Both deletion and insertion
are the results of DNA recombination. However, deletion maybe caused
by intramolecular recombination, which may occur more easily than
insertion caused by intermolecular recombination.
Restriction Analysis of Nuclear DNA
The genus Oryza has been subjected to morphometric, cytogenetic,
isozyme, and chloroplast DNA restriction analysis and all have
contributed in complementary ways to our understanding of the genus.
Restriction fragment length polymorphism (RFLP) analysis of nuclear
DNA is a novel tool for studying genetic variation and phylogenetic
relationship among populations and species of Oryza. Since RFLP
analysis is done directly at the DNA level, it expresses the heritable
change in the nucleotide sequence both in coding and non-coding
regions, RFLP studies arc more sensitive to genetic changes than those
of isozymes, which reflect only those changes resulting in specific amino
acid substitutions. Moreover, the available RFLP markers in rice are
more than isozyme markers. It is therefore possible to study a relatively
large number of loci scattered throughout the genome by use of this
technique.
Wang et al. (1991) undertook a study to determine the level of RFLP
variation both within and between the species in the genus Oryza and to
determine the phylogenetic relationship among the species in this genus.
They took 3 to 5 individuals as samples from each of 22 accessions
representing species from the genus Oryza. Each sample was tested with

S.D. Sharma et al. 335
15 RFLP probes using a single restriction enzyme, Eco Rl. They observed
no significant differences between diploids and tetraploids with respect
to within-species-polymorphism. The tetraploid species O, latifolia was
among the highly polymorphic species, suggesting that this species may
not be of recent origin. In general, recently created polyploids are likely
to have limited genetic variation due to the genetic bottleneck imposed
by the polyploidization.event.
Classification of Oiyza species based on RFLP matched remarkably
well with the previous classifications based on morphology, genomic
constitution, and isozymes. Four species complexes could be identified
corresponding to those proposed by Vaughan (1989): the O. ridley
complex, the O. meyeriana complex, the O, officinalis complex, and the O.
sativa complex. Within the O. sativa complex, accessions of O. rufipogon
from Asia (including O. nivara) and perennial forms of O. rufipogon
Australia clustered together with the accessions of cultivated rice O.
sativa. Surprisingly, indica and japónica showed a closer affinity with
different accessions of wild O, mfipogon than with each other,
supporting a hypothesis of independent domestication for these two
types of rice. Australian annual wild rice, O. meridionalis, was clearly
distinct from all other accessions of O. rufipogon, and was considered a
separate species (Ng et al., 1981).
Using genetic relatedness as a criterion, it was possible to identify
the closest living diploid relatives of the currently known tetraploid rice
species. Results from these analyses suggest that the BBCC tetraploids
(O. malampuzhaensis, O. schweinfurthiana, and O. minuta) are either of
independent origin or have experienced "introgression from sympatric
C-genome diploid rice species" (Ng et al., 1981). The CCDD tetraploids
species of America (O. latifolia, O. alta, and O. grandiglumis) may be of
ancient origin since they show a closer affinity to each other than to any
known diploid species. Their closest living diploid relatives belong to C
genome (O. eichingeri) and E genome (O. australiensis) species.
Jena and Kochert (1991) used RFLP analysis to detect genomic DNA
variation within the accessions of O. latifolia, and among the accessions
of O. latifolia, O. alta and O. grandiglumis—the alióte trap loid species
with genomic constitution CCDD. Based on RFLP data, they observed
all these populations to be of a single species rather than of three different
types. This finding was consistent with the reports based on their
morphology, crossability, cytology, isozyme, and chloroplast DNA
analysis (Nezu et al., 1960; Gopalakrishnan and Sampath, 1966;
Katayama, 1967; Jena and Khush, 1984). A phylogenetic tree, based on
parsimony analysis by Jena and Kochert (1991) grouped the accessions
into clusters which fitted with their geographic origins. Based on this,
they conjectured the origin of some of the accessions. For example, their

336 Rice Breeding and Genetics: Research Priorities and Challenges
analysis clustered one accession of O. latifolia with two other accessions
(one of O. alia and one of O. latifolia), which were from Brazil^ suggesting
that the accession O. latifolia might have originated from Brazil only.
RFLP for Intraspecific Differentiation
Ishii etal (1995) carried out an experiment to clarify the nuclear genome
differentiation in Asian varieties of O. sativa based on the restriction
fragment patterns with two endonucleases. Eco RI and Hind III, and
using 12 single-copy rice DNA probes. They found 93 types of nuclear
genomes among 112 local varieties from 17 Asian countries. Those 93
types of nuclear genomes were divided into eight groups, namely. A,
Bl, B2, C l, C2, Dl, D2, and E. These results were compared with the
previous isozyme analysis and RFLP analysis on chloroplast genome
using the same varieties. Classification of isozyme analysis matched
well with that.of nuclear genome indicating synchronous differentiation
of isozyme constitutions and nuclear genome in Asian varieties. Nuclear
genomes were grouped into indica (A, Bl, and B2), intermediate (Cl,
C2, and Dl) and japónica (D2 and E) types. Earlier chloroplast (ct) DNA
was studied by Ishii et al. (1988) and Dally and Second (1990), who
found two major chloroplast genome types (type 1 and type 3) in
japónica and indica varieties. Ishii et al. (1993) also examined
mitochondrial DNA variation O. sflhufl and found that the mitochondrial
genome differentiated between japónica and indica varieties. They concluded
that the japónica group with D2 and E nuclear genomes has only
the type 1 chloroplast genome, whereas indica and[ intermediate groups
contain both (type 1 and type 3) chloroplast genomes. It is noteworthy
that the type 3 chloroplast genome, which was not found in the japónica
group was the dominant type in the indica varieties. The result indicates
that differentiation of the nuclear genome had partially synchronized
with that of the chloroplast genome.
Zheng et al. (1994) studied three indica and three japónica testers for
wide compatibility along with 21 wide-compatibility varieties (WCV)
using 160 RFLP probes and four enzymes: Eco RI, Eco RV, Hind III, and
Xba I. They found that 68 out of 160 probes were indica-japonica tester
differentiated and produced identical hybridization patterns between
subsepecies. Qian et al. (1995) tested those 68 indica-japonica tester
differentiating probes in seven indica and seven japónica varieties to
distinguish subspecies differentiating probes. Twenty-one probes were
confirmed to be subspecies differentiating probes and showed different
hybridization patterns between indica and japónica subspecies with at
least one enzyme digestion. For 19 of the 21 indica-japonica

S.D. Sharma et al. 337
differentiating probes^ polymorphisms between indica and japónica
were detected by more than one enzyme^, indicating that most of the
polymorphisms between the two subspecies were due to insertions/
deletions (McCouch et ah, 1988). This implies that rearrangements in the
genome have played an important role in the evolution of cultivated
rice.
Repetitive Sequences for Genome Specificity
One distinguishing feature of the genomes of most of the higher
eukaryotes is the presence of large amounts of repetitive DNA, In higher
animals, many repetitive sequences are well characterized in terms of
their length, abundance, chromosomal distribution, and even nucleotide
sequences. Recently, a number of studies on repetitive sequences have
been reported in higher plants such as rye, wheat, barley, maize, flax,
and rice. One important conclusion drawn from these studies is that
repetitive sequences appear to change rapidly during evolution and are
useful in studying genome evolution at the molecular level. The rice
nuclear genome contains approximately 50% repetitive DNA as determined
by Q t analysis (Deshphane and Ranjekar, 1980; Zhou, 1986).
Zhao et al. (1989) undertook a study to analyze the divergence of rice
species and the evolutionary relationship among related genomes.
Thirty-seven rice entries were used in their experiment, representing 13
species. They used four repetitive sequences as probes to screen all the
rice DNA samples from the 37 entries by slot-blot hybridization. When
pOs 48 was used as the probe, only DNA samples from the AA genome
showed hybridization. This indicates that the repetitive sequence of pOs
48 is A A genome specific although the copy number in various AA
genome rice varieties differs.
When pOa 4 was used as a probe, strong hybridization was
observed, mainly with O. australiensis DNA (EE genome). Much weaker
hybridization was barely visible with O. alta and O. latijblia (CCDD
genome) but no hybridization was found with other genomes. This
shows that repeated sequencing of pOa 4 is EE genome specific. Weak
hybridization to the CCDD genome suggests that the CCDD and EE
genomes are more closely related to each other; pQa2 only hybridized to
DNA from O. officinaiis (CC genome), indicating that this repetitive
sequence is CC genome specific. It is interesting that this probe did not
hybridize with O. alta and O. latifoHa (CCDD genome). The absence of
this sequence in certain species with a CC complement suggests that this
CC genome specific sequence may be lost in the CCDD genome of O. alta
and O. latifoUa. Alternatively, there may be two subtypes of the CC

338 Rice Breeding and Genetics: Research Priorities and Challenges
genome which contain different repetitive DNAs. The probe pObl is FF
genome specific since it only hybridized with DNA from O. hrachyantha.
In order to determine the specific chromosome location of the rDNA
probes, Cordesse et ai (1992) analyzed the nuclear ribosomal gene
intergenic spaces in rice. This study confirmed that spacer variability in
wild species is due to variation in the copy number of the 250-260 bp
repeats and, furthermore, that other regions of the spacer are also
involved in variability,
Various genomes, as observed by rDNA technology, fall into groups
that roughly correspond to those defined by extensive use of isozyme
variation and RFLP (Second, 1984; Dally and Second, 1990). The AA and
BB genomes are rather close to each other, differing only in the region
spaimed by pRR 217-10 and 11. The CC genome seems to be more
divergent because, although CC spacer fragments hybridized to the
same probes as the BB fragments, the intensity of the hybridizing bands
was always lower. The EE genome shows extensive divergence. The
rDNA spacer from the FF genome shows the closest similarity to that of
AA with only one probe discriminating between them. Since the
distribution area of O. hrachyantha largely overlaps that of other species
of the AA genome in Africa, the possibility exists for a coevolution of
ribosomal genes between the two genomes or an early introgression of
AA genome ribosomal genes into the FF genome.
The other feature revealed by this analysis was the discrepancy
between observation on CC and CCDD genomes. Genes coming from
the CC genome had perhaps been eliminated in the allotetraploid.
Cloning of these genes should provide probe(s) to trace the DD genes in
wild relatives of rice in America as CCDD species are found only in this
continent.
PHYLOGENETIC RELATIONSHIPS AND EVOLUTIONARY
TRENDS
Studies on the comparative morphology of Oryza species and cytogenetic
studies of their interspecific hybrids and genome analysis have
contributed much information regarding the relationship between
species of the genus Oryza. Sharma and Shastry (1971) have discussed
the primitive or advanced nature of various characters and, based on
these, Sharma (1986) presented a comprehensive picture of the
evolutionary trends in the genus Oryza. According to him, the genus
was initially represented by small-size plants growing in well-drained
moist soil in equatorial forests under high humidity conditions. The
primitive species were small in stature and adapted to forest shade.

S.D. Sharma et a h 339
Subsequently a gradual evolution in ihe genus from forest shade to open
habitat took place and, in this process, the species became hydrophytes
to maintain their physiological homeostasis. Thereby the species of
Oryza became robust, tall, with well-ramified panicle and even larger
size spikelets.
Sect, Padia represents the most primitive group of species (Sharma,
1986). O. schlechteri and O, meyeriana are small-size plants growing in
well-drained soils in forest shade. Sect. Angustifolia occupies a relatively
advanced position in the evolution of the genus Oryza. Of these, O.
perrieri and O. tisseranti are the perennial species and more primitive
than O. hrachyantha and O. angustifolia. All these species have retained
the small plant size but have adapted to open habitat and aquatic
conditions.
The large plant size does not manifest itself in the genus until the
evolution of Sect. Oryza. In this section, Ser. Latifoliae is comparatively
primitive, with most of the species adapted to partial shade and moist
soils, near running streams, etc. In some species, such as O. eichingeri,
the plant and panicle size are comparatively small and the species grows
in well-drained moist soil in the humid tropical forests of Uganda and
Sri Lanka. On the other hand, O. officinalis and O. latifolia are
represented by much larger plant size with well-ramified panicle. These
species are still adapted to partial shade and are only partly hydrophytes.
O. australiensis, on the other hand, is adapted to open habitat and
grows in pools of water. In Ser. Sativae, the plants are completely
hydrophytic and adapted to open habitat. The major trend in this series
has been from perennial habit to annual habit, gregarious habit, and
greater seed production. A detailed description of the phylogenetic
trends in each of these cases is presented below.
Sect. Padia
A compartiave study of species in Sect. Padia indicated O. schlechteri to
be primitive. It is a diploid {2n = 24), has retained a smooth surface of
fertile lemma and palea with absence of setae, and lacks awns. However,
it occupies an advanced position by virtue of reduction in the spikelet
size and setiform sterile lemmas, which may occasionally even be
missing. Therefore, it may be considered an early offshoot arising from
the common stock that gave rise to other species of this section.
Compared with O. ridleyi, O. meyeriana is more primitive as it is a
diploid (2n = 24). Moreover, the plants are rhizomatous and the spikelets
are awnless. On the other hand, the sculpturing of fertile lemma and
palea and setiform sterile lemmas could be viewed as advanced
characters. The transverse section of fertile lemma indicates a thick band

340 Rice Breeding and Genetics: Research Priorities and Challenges
of sclerenchymatous tissue that is absent in O. ridleyi (Sharma^ 1964), All
these characters indicate that O. meyeriana, though primitive in the
section, is yet highly specialized in its own direction.
O. ridleyi (including O. longiglumis) is represented by comparatively
robust plants not met within the Sect. Padia outside Set, Ridleyanae. The
panicle morphology is also different from that of O. meyeriana (sensu
lato) and O. schlechteri. The spikelets are characterized by long glumes
and awns not met with in the latter two species. The amphidiploid
nature of the species indicates that it has evolved through hybridization
of two different species possessing dissimilar genomes.
O. meyeriana and O. ridleyi share many characters. Both species have
slender culms, thin and herbaceous leaves, absence of pulvinus at the
base of branches of the panicle, termination of the axis of the spikelets
with pedicel, straight rachilla, setaceous sterile lemma, sparse hairiness
of fertile lemma, and long cylindrical caryopsis. The two species are
sympatric and prefer shady habitat. Roy (1966) has pointed out many
similarities between O. meyeriana and 0. ridleyi in the anatomy of stem
and leaf as well as in the arrangement of marginal hairs in the leaf blade.
Tateoka (1963) studied the embryo structure in the genus Oryza and
found that O. meyeriana shows characters similar to those of O. ridleyi. It
is, thus probable that one of the genomes of O. ridleyi could be the same
as that of O, meyeriana (Sampath, 1962).
Sect. Angustifolia
O, perrieri exhibits primitive characters such as perennial habit,
nonmucronate fertile lemma and slender awns. O. tisseranti also shares
these characters. Besides, it is a rhizomatous species. On the other hand,
O. hrachyantha expresses some of the advanced traits such as oblique
articulation of the pedicel, comma-shaped rachilla, mucronate fertile
lemma and robust awns. O. angustifolia closely resembles O. hrachyantha.
O. perrieri and O. tisseranti, therefore, represent a primitive series in Sect.
Angustifolia compared to O. hrachyantha and O. angustifolia, which are
advanced.
Sect. Oryza
Ser. Latifoliae\ Five species of Ser. Latifoliae occur in Asia. Of these, three
{eichingeri, officinalis, rhizomatis) are diploid having CC and DD genomes
and two {minuta, malampuzhaensis) are tetraploid having the BBCC
genome. Only one species {officinalis) has a wide distribution spreading
from the west coast of India to the Philippines and New Guinea. The

S.D. Sharma ei al. 341
other four species are confined to small localities only. The African
elements of Ser. Latifoliae are represented by two diploid species
{punctata, eichingeri) and one tetraploid species (schweinfurthiana). The
two diploid species represent BB and CC genomes and the tetraploid
species represents the BBCC genome. The diploids {punctata, eichingeri)
are confined to small areas in east and central Africa respectively and
the tetraploid {schweinfurthiana) is distributed widely frorr\ the Ivory
Coast to Madagascar in that continent. In America^ three species of Ser.
Latifoliae are found and all of them are tetraploid^ representing, the
CCDD genome. Of these, O. grandiglumis has a central distribution
whereas O. latifoUa is distributed widely, spreading from the West
Indies and southern Mexico up to Peru, Bolivia, and Paraguay. O. alta
occupies an intermediate distribution.
It is remarkable that in Asia, a diploid species {ojficinalis) is more
successful in wider distribution than the tetraploids {minuta, malampuzhaensis).
On the contrary, in Africa, the diploid species {eichingeri,
punctata) have restricted distribution and a tetraploid {schweinfurthiana)
species is distributed extensively. In America, the diploids are missing
and the three tetraploids are widely distributed.
Ser. Sativae: The Asiatic plants of Ser. Sativae are represented by
three species, viz. O. rufipogon,.0. nivara, and O, sativa. Of these, the first
is a perennial species whereas the other two are annual. Various authors
(Porteres, 1950; Richharia, 1960; Morishima et ah, 1961; Sampath, 1962)
have suggested that the perermial species is the most primitive in this
group. The origin of an annual species from a perennial one represents a
natural sequence of evolution. This is also true of the other morphological
characters that differentiate O. nivara from O. rufipogon (Sharma and
Shastry, 1965a, 1965b). The geographical distribution of O. nivara
(including that of O. meridionalis) is more or less peripheral to that of O.
rufipogon. The wider distribution of O. nivara {including O. meridionalis)
especially in the fringes of O. rufipogon is explained if O. nivara is
assumed to have evolved from O. rufipogon through mutation and
natural selection in the geological past.
The origin of the Asian cultivated rice, O. sativa, has been discussed
in detail in a separate chapter. The origin of the African cultivated rice,
O. glaberrima, has been discussed in detail by Porteres (1956). According
to him, this cultivated rice originated in the montane regions of Senegal.
Secondary centers of origin developed around the Sokoto River in
Nigeria and Lake Chad. According to Porteres (1956) and Mishra and
Misro (1969), the African species have been somewhat differentiated
into two groups, termed by them as japonicoides and indicoides.
Morishima et al (1962) could not notice any such differentiation. But this
differentiation of O. glaberrima is not as accentuated as that of O. sativa,

342 Rice Breeding and Genetics: Research Priorities and Challenges
probably because the O, glaberrima did not get variation in altitude as the
O, sativa had to come across during its evolution.
References
A ndrew s, F. W. 1956, The Flowering Plants of the Sudan, voL III. T. Buncle & Co. Ltd, A rbroath,
Scotland.
Bisw al, J. and Sharm a, S.D . 1987. Taxonom y and phylogeny oiOryza collina. Oryza 2 4 :24r-29.
Boorham ont, J. 1962. Recherches cytogenetigues chez guelgues hybrids interspecifigues d'
Oryza Cellule 63(1): 52-132.
Cam us, A. 1926. Y vesia, genre nouveau et espèces nouvelles de Gram iness m alagaches. Bull
Sec. BoLFr. 73: 687-691.
Chang, T.T.. 1985 Crop history and genetic conservation: Rice-a case study. Iowa State}. Res.
59 (4): 425-456.
Chaterjee, D. 1948. A m odified key and num eration of the species of genus Oryza Linn.
Indian /. Agric. Set. 8 :1 8 5 -1 9 2 .
Chevalier, A. 1932. N ouvelle contribution a Tetude system ateque des Oryza. Rev. Bot. Appl.
Agric. Trop. 12; 1014-1032.
Clayton, W .D. 1968. Studies in Gram ineae. XVII. Kew. Bull 21(3): 485-488.
Cordesse, F., G rellet, F., Reddy, A.S. and D elseny, M. 1992. Genom e specificity of rDNA
spacer fragm ents from Oryza sativa L. Theor. Appl. Genet. 83: 864-870.
Dally, A.M . and Second G. 1990. Chlorpolast D N A diversity in w ild and cultivated species
of rice (Genus Oryza, Section Oryza)'. C ladistic-m utation and genetic distance analysis.
Theor. Appl. Genet. 80: 209-222.
Deshphane, V.G. and Ranjekar, P.K. 1980. Repetitive D N A in three G ram inaeae species with
low DNA content. Hoppe Seyler's Z Physiol Chem. 361:1223-1233.
Dhua, S.R. 1994. Genom e analysis of Oryza rhizomatis Vaughn. Ph.D. Thesis. V isva Bharati,
Sriniketan, India.
Ellis, J.L. 1985. Oryza indandamanica Ellis, a new rice plant from islands of Andam ans. Bull
Bot. Surv, India 27; 225-227.
Fukui, K., O hm ido, N. and Khush, G,S. 1994. Variability in rDNA loci in the genus Oryza
detected through flourescence in situ hybridisation. Theor. Appl Genet, 87; 893-899.
G opalakrishnan, R. 1959.. Cytogenetical studies on interspecific hybrids in the genus Oryza.
A ssociateship thesis, Indian Agric. Res. Inst., New Delhi.
G opalakrishnan, R. 1964. Spéciation in Ser. Latifoliae. Unpub. Ph.D. thesis, Indian Agric.
Res. Inst. New Delhi.
G opalakrishnan, R. 1965. Subspeciation in Oryza officinalis W all, ex Watt./. Biol. Sci. 8(2): 6 0 -
67.
G opalakrishnan, R. 1966. Taxonom ic status of Oryza collina (Trim en) Sharm a et Shastry.
Indian }. Genet. 26(1): 98-100.
Gopalakrishnan, R. and Sam path, S., 1966, The Am erican species of Oryza. Oryza 3 (1): 3 5 -
40.
G opalakrishnan, R. and Sam path, S. 1967. Taxonom ic status and origin of A m erican
tetraploid species of the series Latifoliae Teteoka in the genus Orym. Indian /. Agric, Sci.
37: 465-475.
Gopalakrishnan, R., Sharm a, S.D. and Shastry, S.V.S. 1965. Genom e cortstitution of Oryza
iiifl/amp/iz/wcMesis Krishn. et Chandr. Cwrr, Sd. 34 (4): 128.

S.D, Sharma et at. 343
G otoh, K. and O kura, E. 1935. Cytogenetical and genetical studies of Orym. jjm .}. Genet. 11
(2 ): 130-131 (in Japanese).
Grover, A. and Pental, D. 1992. Interrelationships o f Oryza species based on electrophoretic
patterns of alcohol dehydrogenase. Can. /. Bot. 7 0 :3 5 2 -3 5 8 .
H irai, A., Ishibashi, T., M orikam i, A., Iwatsuki, N ., Shinozaki, K. and Sugiura, M . 1985. Rice
chroroplast D N A; a physical map and the location of the genes for the large subunit of
ribulose 1, 5-biphosphate carboxylase and the 32-KDa photosystem II reaction centre
protein. Theor. Appl, Genet. 7 0 :117-122.
H iratsuka, J., Shim ada, H,, W hittier, R., Ishibashi, T., Sakam oto, M ,, M ori, M ., Kondo, C.,
H onji, Y ., Sun, C .R., M eng, B.Y., Li, Y.Q., K anno, A., N ishizaw a, Y., H irai, A., Shinozaki,
K. and Sugiura, M . 1989, The com plete sequence of the rice {Oryza sativa) chloroplast
genom e; inter m olecular recom bination betw een distinct tRNA genes accounts for a
m ajor plastic DNA inversion during evolution of the cereals. Mol, Gen. Genet. 2 1 7 :1 8 5 -
194.
H irayoshi, T. 1937. Species hybrids betw een cultivated rice and w ild rice {Oryza latifoUa)
with special reference to their sterility and m aturation division (a prelim inary note)./pM.
/. Genet. 13: 5 9 -6 0 (in Japanese).
Hu, C.H. 1967, Cytogenetic studies oi Oryza officinalis com plex, II. M eiotic studies of induced
autotreraploids of O, ojficinalis. Bot. Bull. Acad. Sinica 8 (spec, no); 327-338.
H u, C.H . 1970. C ytogenetic studies of O. officinalis com plex. III. The genom ic constitution of
O. punctata and 0 . eichingeri. Cytologia 35 (2): 304-318.
H ubbard, C.E. 1950. Oryza angustifolia C.E. Hubbard, Hooker's Incones Plantarum (Fifth Series),
v o l.V . Tabula 3492.
Ichikaw a, H., H irai, A. and Katayam , T, 1986. G enetic analysis of Oryza species by m olecular
markers for chloroplast genom es. Theor. Appl. Genet. 72: 353-358.
IRRI. 1964a. Rice Genetics and Cytogenetics, Elsevier, A m sterdam . 274 pp.
Ishii, J., Brar, D .S., Second, G., Tsunew aki, K. and Khush, G.S. 1995,. N uclear genom e
differentiation in A isan cultivated rice as revealed b y RFLP analysis, Jpn. /. Genet. 70:
643-652.
Ishii, T., Terachi, T. and Tsunew aki, K. 1986. Restriction endonuclease analysis o f chloroplast
D N A from cultivated rice species, Oryza sativa and O.glaberrima fpn. J. Genet. 6 1 :5 3 7 -5 4 1 .
Ishii, T., Terachi, T. and Tsunew aki, K. 1988. Restriction endoiiuclease analysis o f chloroplast
D N A from A -genom e diploid species of Oryza. Jpn. J. Genet. 6 3 :5 2 3 -5 3 6 .
Ishii, T., Terachi, T., M ori, N. and Tsunew aki, K. 1993. Com parative study of the chloroplast,
m itrochondrial and nuclear genom e differentiation in two cultivated rice species, Oryza
sativa and O. glaberrima by RFLP analysis. Theor. Appl. Genet. 86: 88-96.
Jena, K.K. and Khush, G .S. 1984. Em bryo rescue and its scope in rice im provem ent. Rice
Genet, Newslett. 1:133.
Jena, K.K. and Kochert, G. 1991.. Restriction fragm ent length polym orphism analysis of
C C D D genom e species of the genus Oryza L. Plant Molec. Biol. 1 6 :8 3 1 -8 3 9 .
K anno, A. and H irai, A. 1992. Com parative studies of the structure of chlorplast DNA from
four species o f Oryza: Cloning and physical maps. Theor. Appl. Genet. 83: 791-798.
Katayam a, T. 1996 a, C ytogenetic studies on the genus Oryza, 2. Chrom osom e pairing in the
interspecific hybrid w ith the A BC genom es. Jpn. J, Genet. 41(4): 307-316.
Katayam a, T. 1966 b, C ytogenetic studies on the genus Oryza, 3. Chrom osom e pairing in the
interspecific hybrid with the A BC genom es. Jpn. J. Genet. 4 1 (4 ); 317-324.
Katayam a, T. 1967. C ytogenetical studies on Oryza P j hybrids of the crosses BBC C x CC,
BBC C X a diploid strain of O. pimcfflfii and CC x a diploid strain o f O. punctata. Proc. Jpn.
Acfld,43 (4): 327-331.

344 Rice Breeding and Genetics; Research Priorities and Challenges
K atayam a, T.C ., A kiham a, T. and W eliw ita, S.M .P, 1972. D istribu tion and som e
m orphological characters of w ild rice in Ceylon. P relim inary R eport of T ottori
U niversity's Scientific Survey (1971) 1 : 60-64.
Kihara, H., Nezu^ M ., Katayama^ T.C ., M atsum ura, S. and M abuchi, T. 1961. Genom e analysis
in the genus Oryza, II. Ann. Kept. Natl. Inst. Genet. (Japan) 1 1 :4 0 -4 1 .
Kishim a, Y,, M ikam ii, T ., H irai, A., Sugiura, M. and Kinoshita, T. 1987. Beta chlorosplast
genom es; analysis of fraction I protein and chloroplast DNA variation. Theor. Appl. Genet.
73; 330-336.
Laim ert, E. 1965. A survey of the genus Leersia in A frica {Gramineae, Oryzoideae, Oryzeae).
Senck. Biol. 4 6 :1 2 9 -1 5 3 .
Li, H.W . 1964. Studies on genetic and cytogenetic evidence for species relationship in the
Republic of China. In: Rice Genetics and Cytogenetics, Elsevier, A m sterdam , pp. 118-13.
Li, H .W , W eng, T.S., Chen, C.C. and W ang, W .H. 1961, Cytogenetical studies of Oryza sativa
L. and its related species, 1. Hybrids O. pamguaiensis W edd. x O. brachyantha Chev. et
Roehr., O. paraguaiensis W edd. x O. australiensis D om in and O. australiensis D om in x O.
alta Sw allen. Bot. Bull, Acad. Sínica 2; 79-86,
Li, H. W ., W eng, T.S., Chen, C.C. and W ang, W .H. 1962. C ytogenetical studies of Oryza sativa
L. and its related species, 2. A. prelim inary note on the interspecific hybrids w ithin the
Section Sativa Roshev. Bot. Bull. Acad. Sínica, 3 :2 0 9 -2 1 9 .
Li, H .W ., Chen, C.C., W eng, T.S. and W uu, K.D. 1963. C ytogenetical studies o f Oryza sativa L.
and its related species, IV. Interspecific crosses involving O. australiensis w ith O. sativa
and O. minuta. Bot. Bull, Acad. Sínica 4; 65-74.
Li, H .W ., Chen, C.C., Katherine, C.L., Lu, H.K. and Hu, C.H. 1964. Pachytene studies o f the
hybrid Oryza sativa x Oryza officinalis. In; Rice Genetics and Cytogenetics. Elsevier
A m sterdam pp. 141-142.
M cCouch, S.R., Kochert, G., Yu, Z.H., W ang, Z.Y,, Khush, G.S. Coffm an, W .R. and Tanksley,
S.D. 1998. M olecular m apping of rice chrom osom es. Theor. Appl. Genet. 76; 815-829,
M ishra, P.K. and M isro, B.1969. Postulation of two subspecies in the A frican cultivated rice
{Oryzae glaberriona Steud). Indian J. Agrie. Sci. 39 (10): 966-970.
M orinaga, T. 1940. C ytogenetical studies on Oryza sativa L. IV. T he cytogenetics o f F| hybrid
of O. sativa L. and O. minuta Pres. Jpn.). Bot. 1 1 :1 -1 6 .
M orinaga, T. 1941. C ytogenetical studies on Oryza sativa L., V. The cytogenetics o f F j hybird
of O. sativa L. x O. latifolia Desv. )pn. J. Bot. 11:4 6 1 -4 7 8 .
M orinaga, T. 1943. C ytogenetical studies on Oryza sativa, VI. The cytogenetics o f F j hybrid of
O. minuta Presl. and O. latifolia Desv. Jpn. f. Bot. 12:3 4 7 -3 5 7 .
M orinaga, T. 1959. N ote on genom e analysis inOryza species. Inti. Rice Comm. Newslett. 8(3):
10- 11.
M orinaga, T. 1964. Cytogenetical investigation on Oryza species. In; Rice Genetics and
cytogenetics. Elsevier A m sterdam , pp. 91-102.
M orinaga, T. and Kuriyam a, H. 1956. C ytogenetical studies on Oryza sativa L., VIII. The F^
hybrids o f O. sativa L. and O. cubensis Ekm an. Jpn. J. Breed. 6; 133-141.
M orinaga, T. and Kuriyam a, H. 1957. O n the interspecific hybrid of O, sativa and O.
breviligulata. Jpn. J. Breed (Suppl.) 7:12 (in Japanese).
M orinaga, T. and Kuriyam a, H. 1959. Genom ic constitution of Oryza officinalis. Jpn J. Breed. 9;
259.
M orinaga, T. and Kuriyam a, H. 1960. Interspecific hybrids and genom ic constitution of
various species in the genus Oryza. Agrie. Hortic. 35 (5 -8 ): 77B-776,9 3 5 -9 3 8 ,1 0 9 1 -1 0 9 4 ,
1245-1247 (in Japanese).
M orinaga, T., Kuriyam a, H. and O no, S. 1960. O n the interspecific hybrid of Oryiui minuta
and O. australiensis. (Abs.) Jpn. J. Genet. 3 5 :2 7 8 (in Japanese).

S.D. Sharma et al. 345
Morinaga, T., Kuriyama, H. and Ono, S. 1962. Oryza grandiglumts and its interspecific
hybrids. (Abs.) Jpn. /. Breed. 12 : 61. (in Japanese).
Morishima, H. 1984. Wild plant and domestication. In: Biology of Rice, S. Tsunoda and N.
Takahashi (eds.). Elsevier, Amsterdam pp. 3~30.
Morishima, H., Oka, H.L and Chang, W.T. 1961. Direction of differentiation in populations
of wild rice O. perennis and O. sativa f. spont. Evolution 15; 326-339.
Morishima, H., Hinata, K. and Oka, H.1.1962. Comparison between two cultivated rice, 0.
sativa L. and O. glaberrima Steud. Jpn. J. Breed. 12(3): 153-165.
Nandi, H.K. 1938. Interspecific hybridisation of Oryza, I. Cytogenetical evidence of the
hybrid origin of Oryza minuta Presl. Trans. Bose Res. Inst., Calcutta 11 (1935-36) : 99-121,
Nezu, M., Katayama, T.C. and Kihara, H. 1960. Genetic study of genus Oryza, I. Crossability
and chromosomal affinity among 17 species. Seiken Ziho 11:1-11.
Ng, N.Q., Hawkes, J.G., William, J.T. and Chang, T.T. 1981. The recognition of a new species
of rice {Oryza) from Australia. Bot.}. Linn. Soc. 82:327-330.
Nowick, E.M. 1986. Chromosome pairing in Oryza satim L. x O, latifolia Desv. hybrids.
Canadian J. Genet, Cytol. 28(2): 278-281.
Ogihara, Y. and Tsunewaki, K. 1988. Diversity and evolution of chloroplast DNA inTriticum
and Aegilops as revealed by restriction fragment analysis, Tfieor. Appl. Genet. 76:321^32.
Pilger, R. 1914, Neu und weniger bekannte Gramineen aus Papuasien. Botan. Jahrb. 52:167-
176.
Porteras, R. 1950. Primitive at progrès dans l'évolution a sein due genre Oryza. Rev. Bot.
Appl. Agric. Trop. 30: 604-610.
Porteras, R. 1956. Taxonomie agrobotanigue dez riz cultives: O. sativa Linné et O.glaherrina
Steudel. J. Agric. Trop. Bot. Appl. 3; 341-384,541-580,627-700,821-856.
Prodoehl, A. 1922. Oryzeae monographice discribuntur B of. Arch. 1:211-224,231-256.
Qian, H.R., Zhuang J.Y., Lin, H.X., Lu, J. and Zheng, K.L. 1995. Identification of a set of RFLP
probes for subspecies differentiation in Oryza sativa L. Theor. Appl. Genet. 90; 878-884.
Ramanujan, S. 1937. Cytogenetical studies in the Oryzae, III. Cytogenetical behaviour of an
interspecific hybrid in Oryza. J. Genet. 35:223-258.
Richard, B, Moequot, B. and Fournier, A„ et al, 1986. Expression of alcohol dehydrogenease
in rice embryos under anoxia Plant Mol. Biol. 7:321-329.
Richharia, R.H, 1960. Origin of cultivated rices. Indian }. Genet. 20 (1); 1-14.
Roschevicz, R.J. 1931. A contribution to the knowledge of rice. Appl. Bot. Genet. PI. Breed. Bull.
27 (4): 3-133.
Roy, J.K. 1966. Anatomical studies in genus Oryza. V. Interrelationship of some Oryza
species as shown by anatomical characters. Oryza 3(1): 86-95.
Sampath, S. 1962. The genus Oryza: its taxonomy and species relationship Oryza 1(1): 1-29.
Sampath, S. and Subramanyam, M.D. 1968. The taxonomy of a wild Orym species from
Ceylon. Oryza-5 (1): 75-76,
Second, G. 1982. Origin of the genic diversity of cultivated rice {Oryza spp.); Study of the
polymorphism scored at 40.isozyme loci, Jpn. J. Genet. 57:25-57.
Second, G. 1984. A new insight into the genome differentiation in Oryza L. through isozyme
studies. In: Advances in Chromosome and Cell Genetics. A.K. Sharma and A. Sharma, (eds.).
Oxford and IBH Press, New Delhi, pp, 45-78.
Shama Rao, H.K. and Seetharaman, R. 1955. An interspecific hybrid in Oryza. Curr. Sei. 24
346-347.
Sharma, S.D. 1964. Interspecific relationships in genus Oryza. Unpuble. Ph.D. Thesis, Indian
Agricultural Research Institute, New Delhi.

346 Rice Breeding and Genetics: Research Priorities and Challenges
Sharma, S.D. 1986. Evolutionary trends in genus Oryza. In; Rice Genetics, IRRI, Manila,
Philippines, pp. 59-67.
Sharma, S.D. and Sampath, S, 1985. Taxonomy and species realtionship. In; Rice Research iw
India, KAR, New Delhi, pp. 21-43.
Sharma, S.D. and Shastry, S.V.S. 1965a. Taxonomic studies in genus Oryza, III. O. rufipogon
Griff, sensu stricto and O. nivara Sharma et Shastry nom. nov. Indian J. Genet. 25 (2) : 157-
165.
Sharma, S.D. and Shastry, S.V.S. 1965b. Taxonomic stitdies in genus Oryz«, IV. The Ceylonese
Oryzrt spp. affin. O, Wall, ex Watt- Indian J. Genet. 25 (2) ; 168-172.
Sharma, S.D. and Shastry, S.V.S. 1965c. Taxonomic studies in genus Oryza, VI. A modified
classification of the genus. Indian /. Genet. 25 (2): 173-178.
Sharma, S.D. and Shastry, S.V.S. 1966. Taxonomic studies in genus Oryza. V. Sclerophyllum
coearctatum (Roxb) Griff. Bull. Bot. Surv. India 8:42-44.
Sharma, S.D., Subba Rao, G. and Rao, V.J.M., Cytology of Oryza officinalis (4n) x O. latifolia.
Curr. Sci. 43:585-586.
Sharma, S.D. and Shastry, S.V.S. 1971. Phylogenetic studies in genus Oryza, I. Primitive and
advanced characters. Riso 20(2): 127-136.
Shastry, S.V.S. 1965. Genomic differentiation in the genus Oryza, Indian J. Genet. 26A: 258-
; 271. .
Shastry, S.V.S., Sharma, S.D. and Ranga Rao, D.R. 1961. Pachytene analysis of Oryza, III.
Meiosis in an intersectional hyrbid, O. sativa x O. officinalis. Nucleus 4 ; 67-80.
Smith, J.S.C. and Smith, O.S, 1990. Fingerprinting crop varieties. Adv. Agron. 47 ; 85-140.
Tateoka, T. 1962a. Taxonomic studies oî Oryza, I. O. latifolia complex. Bof. Mag. 7 5 :418-427.
Tateoka, T. 1962b. Taxonomic studies of Oryza, II. Several species complexes. Bof. Mag, 75:
455-461.
Tateoka, T. 1963. Taxonomic studies of Oryza, III. Key to the species and their enumeration.
Bof. Mag. 76:165-173.
Tateoka, T. 1964. Report of Explorations in East Africa and Madagascar. National Science
Museum, Tokyo 16 pp.
Vaughan, D.A. 1989. The genus Orym L. Current status of taxonomy. IRRI Research paper
Series 138:1-21.
Vaughan, D,A. 1994. The Wild Relatives of Rice: A Genetic Resource Handbook. IRRI, Manila,
Phlippines, 137 pp,
Vaughan, D.A., Feral, P. and Gerogo G. 1991. A preliminary note on Oryza schlechteri. Rice
Genet. Newslett. 8: 75-76.
Wang, Z.Y., Second G. and Tanksely, S.D. 1991. Polymorphism and phylogenetic
relationships among species in the genus Oryza as determined by analysis of nuclear
RFLPs. Theor. Appl. Genet. 83:565-581,
Watanabe, Y. and Ono, S. 1965, Cytogenetic studies on the artificial polyploids in the genus
Oryza 1. Colchicine induced octoploid plants of Oryza latifolia Desv. |pn, J. Breed. 15(3);
149-157,
Watanabe, Y. and Ono, S. 1966, Cytogenetic studies on the artificial polyploids in the genus
Oryza, 2. Colchicine induced octopolyploid of Oryza minuta Presl. Jpn. J. Breed. 16 (4):
220-230.
Yeh, B, and Henderson, M.T. 1961, Cytogenetic relationship between cultivated rice, Oryza
sativa L. and five wild diploid forms of Oryza. Crop, Sci. 1:445-450.
Yeh, B, and Henderson, M.T. 1962, Cytogenetic relationship between African annual diploid
species of Oryza and cultivated rice, O, sativa L. Crop Sci. 2 (6): 463-467,
Zhao, X., Wu, T,, Xie, Y, and Wu, R, 1989. Genome specific repetitive squences in the genus
Oryza. Theor, Appl. Genet, 78; 201-209.

S.D. Sharma et al. 347
Zheng, K., Qian, H., Shen, B., Zhuang, J., Lin, H. and Lu, J, 1994. RFLP based phylogenetic
analysis of wide compatibility varieties in Otyza sativa L. Theor. Appl. Genet. 88:65-69.
Zhou, G.Y. 1986. Distantly related hybridisation and genetic engineering of crops. In: Rice
Genetics. IRRI, Manila, Philippines, pp. 867-876.
Sipecies
O. alta Swallen
O. angustifoUa C.E. Hubbard
O. australiensis Domin
O. barthii A. Chev.
(Syn. O. breviligulata
A. Chev. et Roehr.)
O. hrachyantha A. Chev. et Roehr.
O. eichingeri A. Peter
O, glaherrima Steud.
O. glumaepetula Steud.
(Syn. O. cubensis Ekman)
O. granúlala Nees et
Arn. ex Hook. f.
O. latifoHa Desv,
O. longiglumis Jansen
O. longistaminata A. Chev. et Roehr
O. meridionalis Ng
O. meyeriana
(Zoll, et Mor. ex Steud.)Baill
O. minuta J.S. Presl ex C.B, Presl
O. nivara Sharma et Shas try
O. officinalis Wall, ex Watt
O. perrieri A. Camus
O. punctata Kotschy ex Steud.
O. ridleyi Hook. f.
O. rufipogon Griff.
O. sativa L.
O. schlechten Pilger
O. schweinfurthiana Prod.
O. tisseranti A. Chev.
Appendix
Species of O ryza and their geographic distribution
Distribution
Central and South America
South Africa
Australia
West Africa
West and central Africa
East Africa and Sri Lcinka
Tropical West Africa
Tropical America and West Indies
South and Southeast Asia
Central and South America
New Guinea
Tropical Africa
Tropical Australia
Southeast Asia
The Philippines
South and Southeast Asia
South and Southeast Asia
Madagascar
East Africa
Southeast Asia
South and Southeast Asia
South and Southeast Asia
New Guinea
Tropical Africa
Central Africa

15
Origin of O. sativa and Its
Ecotypes
S,D. Sharma\ Smita Tripathy^ and Jyostnamayee Biswal^
INTRODUCTION
The origin of the Asian cultivated rice {Orym sativa L.) has been a
debated subject ever since De Candolle (1882) opened this topic for
scientific discussion. However, as more and more data and evidence
have accumulated and our understanding of the subject has become
increasingly clear, finer details have come up for discussion. The salient
contributions in this field have come from Watt (1891), Roschevicz (1931),
Ramiah and Ghose (1951), Sampath and Rao (1951), Richharia (I960),
Sampath (1962), Sharma (1964), Oka (1964, 1974, 1988), Shastry and
Sliarma (1973), Chang (1976), and Morishima (1984). The authors of this
paper have tried to present a new as well as comprehensive hypothesis
about the origin of Asian cultivated rice {Oryza sativa) and its ecotypes,
based mainly on their own findings during the last ten years and have
cited others' contributions in support (For citation of authors for the
binomials used in this paper, please see the Appendix).
There are two cultivated species of rice: O. sativa, which was
domesticated in South and Southeast Asia, is now widely cultivated in all
the rice-growing areas of the world, while O. glaberrima, which was
domesticated in tropical west Africa has remained coniined to that part of
^M. S. Swaminathan Research Foundation, Chennai - 600 113
^ Jayadev College, Naharakanta, Bhubaneswar - 752101

350 Rice Breeding and Genetics: Research Priorities and Challenges
the world. In fact, the Asian cultivated rice (O. saliva) is now so widely
cultivated in the homeland of O. glaberrima that the former may edge out
the latter sooner or later. Morphologically, the two cultivated species not
only look very similar but also show parallel variation. Cytogenetically,
these two species are diploid, have the same chromosome number (2n =
24) and also the same genomic constitution (AA), although they differ at
the subgenomic level (Yeh arid Henderson, 1961; IRRI, 1964b). The
hybrids of the two species are highly sterile.
The situation in the two cultivated species is vicarious as each has a
closely related annual wild species and also a perennial wild species.
Besides, each of the two cultivated species hybridizes in nature with its
annual wild relative forming partially fertile hybrids. However, the
situation differs with regard to the relation of the cultivated species with
their perermial wild species. In Asia, the cultivated species (O. saliva)
frequently hybridizes in nature with its perennial wild relative (O.
rufipogon) and forms partially fertile hybrids. In Africa, the cultivated
species (O. glaberrima) rarely hybridizes in nature with its perennial wild
relative (O. longistaminaia) and the hybrids, if made artificially, are
highly sterile.
A perennial wild species allied to O. rufipogon of Asia is widely
distributed in tropical America. This has been treated as a distinct and
different species by many biosystematists (Yeh and Henderson, 1961;
Sharma, 1964; Chang, 1985) and has been referred to as O. cubensis, a
nomen nudum, by Yeh and Henderson (1961) and as O. glumaepetula by
Sharma (1964) and Chang (1985). It has been considered a mere variant of
the Asian perennial species (O, rufipogon) by Tateoka (1962) and
Vaughan (1994). Both these taxa have the same number of chromosomes
{2n = 24) and the same genome (AA) but differ subgenomically (Yeh and
Henderson, 1961; IRRI, 1964b).
It may also be mentioned that during the 1950s and 1960s, many of
the rice biosystematists treated the perennial elements of Asia
{rufipogon), Africa (longistaminaia) and America (glumaepetula) as a single
species and identified it as O. peren,nis Moench following Chevalier (1932)
and Chatter] ee (1948). According to this view, the perennial elements of
the three continents are treated as three subspecies of a single species O,
perennis (IRRI, 1964a). In this paper, the binomial O. perennis is
discarded, as suggested by Tateoka (1962), and the three elements are
treated as three distinct species.
TAXONOMIC DELIMITATION AND NOMENCLATURE
The Asian cultivated rice (O. saliva) and its allied taxa that occur in Asia
present a continuous array of morphological features, so much so that

S.D. Sharma et al. 351
the whole group was termed the O. sativa complex by Tateoka (1962). A
better understanding of this group has been gained slowly during the
last one hundred years. The first taxon of tins group to be recognized
was the cultivated rice named O. sativa by Linnaeus (1753). Early
taxonomists and flora writers (Hooker, 1897) considered these wild
relatives of O. sativa merely its variants. Subsequently, the wild elements
were assigned infraspecific ranks within O. sativa. For example, Prain
(1903) designated them as O. sativa va.vfatua and Roschevicz (1931) as O.
sativa f. spontanea.
Chatterjee (1948) recognized three species—a perennial wild, an
annual wild, and the annual cultivated (O. sativa)— for this complex in
Asia. He identified the perennial wild species as O. perennis Moench and
called the annual wild species provisionally O. sativa L. var. fatua Prain
pending its correct identification. Ramiah and Ghose (1951) followed
Chatterjee (1948) and also recognized three species in this complex—a
perennial wild, an annual wild, and the annual cultivated species. They
referred to these three species as O. perennis, O. fatua, and O. sativa
respectively. Sampath and Rao (1951), however, held the view that the
Asian elements of the O. sativa complex consist of a perennial wild
species (their O. perennis) and the annual , cultivated species (O. sativa)
only. According to them, the annual wild types of this complex are
natural hybrids between the perennial wild species (their O. perennis)
and the armual cultivated rice (O. sativa). They referred to the natural
hybrids as O. sativa var. spontanea. However, their view was, based on
observations of the taxa in coastal Orissa only.
Sharma and Shastry (1965a, 1965b) extensively collected various
forms of these wild rices from a wide region in India, studied their
morphology, single plant progenies, ecology, proximity to the cultivated
rice fields and geographic distribution. They recognized four distinct
elements in the O. sativa complex of India: (a) a perennial wild species O.
rufipogon (Bor, 1960; Tateoka, 1962), (b) an armual wild species named O.
nivara in the absence of a valid name (Sharma and Shastry, 1965b, 1966);
(c) the annual cultivated species 0 . sativa, and (d) products of natural
hybridization between the wild species and the cultivated species. These
natural hybrids were further divided into two subgroups, (i) hybrids
between O. rufipogon and O. sativa, and (ii) hybrids between O. nivara
and O. sativa. The classification of the O. sativa complex of South and
Southeast Asia into three distinct species {rufipogon, nivara, sativa) and
two forms of natural hybrids {sativa x rufipogon, sativa x nivara)
necessitates a reexamination of the morphological characters and
ecological preferences of each of these species.

352 Rice Breeding and Genetics; Research Priorities and Challenges
O. R u fip o g o n
O. rufipogon is a perennial species growing in swamps (stable habitat). It
survives in the drier seasons as clumps due to the presence of sufficient
moisture in the soil and regenerates itself in the monsoon season when
the depth of water rises. The culms branch and subbranch at nodes
piercing through the leaf sheath (extravaginal branching). When the
water is too shallow (less than 15 cm), the culm becomes a runner,
rooting at nodes and spreading horizontally on the ground. In deeper
water, the branches and subbranches remain suspended in water. The
leaves are generally at a right angle to the culm. Panicles are well
exserted and erect when emerging from the surface of water. The
branches of the panicle are open, lax, and may be slightly drooping. The
spikelets are slender, anthers are long, filling the spikelet completely,
and the stigma protrudes, favoring cross-pollination. The leaf sheath,
apiculus, awn, the and stigma are pigmented. A detailed morphological
description of this species is available in Sharma and Shastry (1965b).
O. rufipogon is a photoperiod-sensitive species flowering during
November-December. It grows along the margins of ponds and lakes
and the sides of roads and railway tracks if they are swampy. It is
distributed in the coastal plains of India and also occurs in the lower
basins of the Ganga and the Brahmaputra and their tributaries and
distributaries. Outside India, it is reported to occur in southern China,
Southeast Asia, Indonesia, and New Guinea. Forms similar to O.
rufipogon but having larger leaves, greater ramification of panicles, larger
number of spikelets per panicle, and somewhat larger spikelets are also
available in nature due to introgression of characters from O. sativa into
O. rufipogon.
O. N ivara
Compared with O. rufipogon, O, nivara is shorter in height and
semispreading at the vegetative stage but semierect at maturity. It grows
in shallow seasonal ditches that dry off in summer. The plants are armual
and germinate from self-sown seeds during the rainy season. The new
branches appear from the lower nodes only and grow inside (and parallel
to) the leaf sheath and come out at the point of the collar (intravaginal
branching). The leaves are semiopen and not so drooping as in 0 .
rufipogon. The panicle is poorly exserted or even partly inserted. The
number of primary and secondary branches per panicle is less compared
to O. rufipogon. The rachis and the branches of the panicle are stiffer. The
spikelets are shorter but bolder. The awns are longer and more robust.
The pigmentation in plant parts shows much variation. The leaf sheath.

apiculus^ stigma, and awn may or may not be pigmented. This has been
described in detail by Sharma and Shastry (1965b).
O. nivara is a photoperiod-insensitive species and flowers during
September-October. It is found in small populations in seasonal ditches
in northern India as well as in the Deccan plateau. Its occurrence in
Bangladesh and northeast India is rare. Outside India, it is reported to be
distributed in the plateau regions of Myanmar, Thailand, Cambodia,
Laos, south and southwestern region of mainland China, especially in
Guangxi province and its adjoining areas (Shao et ah, 1986).
O. sativa
S.D. Sharma et al. 353
The special feature of O. sativa is that it has differentiated into several
ecogenetic groups and subgroups, which have been called by Morinaga
(1968) ecospecies and ecotypes. Summarizing the works of earlier
workers and his own studies, he recognized four ecospecies, namely
japónica, javanica, aus, and aman (indica) within this species. The other
types discussed in this paper are: (a) the primitive land races of Jeypore
tract of Orissa (Sampath and Govindaswami, 1958; Oka and Chang
1962) and referred to as southeast Indian hill rices or by their acronym
"seihr'O (b) the japonica-like forms that occur in the sub-Himalayan
region of Nepal, Sikkim, Bhutan, Arunachal Pradesh and southwestern
provinces of mainland China; (c) the hill rices of mainland Southeast Asia
that are closely related to javanica types (Chang, 1986; Glaszmann and
Arraudeau, 1986) and referred to by their acronym "hfmsea"; (d) the
tjereh types of Indonesia that resemble aman types of India, and (e) the
shali ecotype of the Brahmaputra valley that ecologically and
agronomically corresponds with the aman type of Bengal.
It is also remarkable that some of the ecotypes of O. sativa are
photoperiod insensitive while the others are photoperiod sensitive.
Futher, the Asian cultivated rice is not an annual species in the strict sense
as many of its cultivars have the capability to ratoon or regenerate and in
this sense have the capability to perennate.
Natural Hybrids
The genetic barrier between either of the wild species (O. rufipogon and
O. nivara) and the cultivated species (O. sativa) is incomplete. This has
led to introgressive hybridization in both directions and occurrence of all
forms of intergrades in nature. Consequently, the taxonomic distinctness
of the three species in nature is blurred and the whole group appears as a
species complex, which was named the O. sativa complex by Tateoka
(1962).

354 Rice Breeding and Genetics: Research Priorities and Challenges
In the coastal plains of India and in the lower valleys of the Ganga
and the Brahmaputra^ O. sativa gets crossed with O. rufipogon and forms
natural hybrids. In the Deccan plateau O. sativa hybridizes in nature
with O. nivara to form natural hybrids. These natural hybrids have been
referred to as spontanea rices in rice literature. The spontanea rices that
invade the rice fields are products of natural hybridization between O.
sativa and 0 . nivara (in plateau regions) or between O. sativa and O.
rufipogon (in coastal regions) followed by repeated backcrosses with O.
sativa. As a result^ the spontanea rices which grow in the cultivated
fields closely resemble the cultivated rice except for shattering of
spikelets at maturity with or without a few other wild characters such as
black husk^ red kerneb presence of awn, etc. Efforts by farmers to
identify them at the vegetative stage and weed them out from the
cultivated fields have acted as a selection pressure for their closer
resemblance to the cultivated rice (Oka and Chang, 1959).
ORIGIN OF CULTIVATED RICE
Earlier Views
Early rice workers held the view that O, sativa mainly originated from
wild species of the O. sativa complex. However, as taxonomic
delimitation and nomenclature of the elements of this complex were not
clear, various authors (see reference this paper) adopted different
binomials. However, they held the view that, besides species of the O.
sativa complex, some other wild species might also have played a role in
the origin of some of the cultivars of O. sativa. In other words, they
assumed that the Asian cultivated rice had a polyphyletic origin. Among
the various putative progenitors, O. officinalis has received the most
serious consideration as this species has well-ramified panicles, high
number of spikelets per panicle, and small-size grains— characters not
met with in either O. nivara or O. rufipogon but present in O. sativa.
Father more, the distribution of O. officinalis is sympatric with that of O.
sativa. However, the two species are ecologically isolated and hence do
not hybridize in nature. Besides, the synthesized Fj hybrids are highly
sterile. Although both the species are diploid (2n = 24), their
chromosomes either do not pair during meiosis (Ramanujam, 1938) or
pair but separate out before metaphase-I without forming chiasmata
(Shastry et al., 1961). The genomic constitution of the two species differs
(O. sativa = A A, O. officinalis = CC or DD?). Any role of O. officinalis in
the origin of O. sativa, therefore, seems improbable. Because of
morphological similarities between O. officinalis and O. minuta, the latter
was also assumed to have played a role in the origin of O. sativa.

S.D. Sharma et a l 355
However^ as O. minuta is a tetraploid species (2n = 48) with a different
genomic constitution (BBCC) occurring only in the Philippines^ it could
not have played any role in the origin of O, sativa.
Another species considered to have played some role in the origin of
O. sativa was Porteresia coarctata. Until 1965, it was treated as a species of
the genus Oryza only and was known as Oryza coarctata. It grows in the
tidal swamps near sea coasts of South Asia and was assumed to have
played a role in the origin of salinity tolerant cultivars of O, sativa. Its
role in the origin of O. sativa was ruled out when it was found to be a
tetraploid (2n = 48) species.
Chatterjee (1951) presented a classical view on the origin of
cultivated rice. According to him, the annual wild species (our O, nivara)
played the major role in the origin of cultivated rice though he did not
rule out the role of O. officinalis. Ramiah and Ghose (1951) recognized
three species in the O. sativa complex of Asia, namely, a perennial wild
species (their O. perennis Moench), an annual wild species (their O.fatua
Koenig), and the cultivated rice (O. sativa) following Chatterjee (1951).
According to them, the aimual wild species is the progenitor of the
cultivated species. Ramiah and Ghose (1951) were the first rice scientists
to attract the attention of other rice scientists to the Jeypore tract of
Orissa as "this area might form another independent center of origin"
(Ramiah, 1953).
Sampath and Rao (1951) treated the perennial wild species of Asia
(rufipogon), Africa {longistaminata), and America (glumaepetula) a s ' a
single species and called it O. perennis Moench, as suggested earlier by
Chevalier (1932) and Chatterjee (1948). They proposed that the peretmial
form of Africa (our longistaminata) has given rise to O. glaberrima in
tropical west Africa and that of Asia (our rufipogon) has given rise to O.
sativa in South and Southeast Asia. According to them, each of the two
forms of O. perennis hybridize in nature with their cultivated
counterparts (O. sativa in Asia and O. glaberrima in Africa) to form
natural hybrids. Tn other words, Sampath and Rao (1951) proposed a
monophyletic origin for cultivated rices of Africa as well as Asia, which
was elaborated by Richharia (1960) and Sampath (1962). The perennis
hypothesis was supported by Oka (1964,1974,1988) who amassed further
evidence of natural hybridization between O. rufipogon (their O. perennis)
and the cultivated rice (O. sativa) in Asia. However, he did not support
Sampath's view that the perennial form of Africa [longistaminata) has
given rise to O. glaberrima in Africa or the view that these two species
frequently hybridize in nature to produce hybrid populations. Sampath
(1962) himself recognized the Asian (rufipogon) and the African
(longistaminata) perennial rices as two distinct and different species and
in this sense demolished his own hypothesis of monophyletic origin of
f i

356 Rice Breeding and Genetics; Research Priorities and Challenges
cultivated rices. Later, Sampath (1962, 1964b) recognized the existence
of an annual wild species in South and Southeast Asia but defined it as
the fixed forms of natural hybrids between the perennial species (his
perennis) and the cultivated species {sativa). Lately, Morishima (1984)
has partially modified the perennis hypothesis and has suggested that
forms intermediate between the perennial and the annual types of wild
rices might have given rise to the cultivated rices.
The Nivara Hypothesis
Sampath (1962) proposed the perennial species (our O. rufipogon) as the
progenitor. This was based mainly on his field observations in coastal
Orissa, where the armual species {nivara) does not occur and all the
annual wild forms of the O. sativa complex occurring in this area are the
product of natural hybridization between O. rufipogon and O. sativa.
However, the recognition of a distinct and different annual species well
distributed over the vast plateau regions of South and Southeast Asia led
Sharma (1964) and Shastry and Sharma (1973) to propose that the
cultivated rice (sativa) of Asia originated from the annual wild species
(nivara). According to them, perennial wild to annual wild and anniual
wild to annual cultivated must have been the natural and logical
sequence of evolution and, therefore, the annual wild species must have
been the progeny (and not the progenitor) of the perennial wild species.
Sharma (1964) and Shastry and Sharma (1973) resurrected the views of
Ramiah and Ghose (1951) with additional evidence, precise taxonomic
delimitations, and valid nomenclature (Sharma and Shastry, 1965a,
1965b, 1966a).
O, nivara is an annual species which grows in shallow seasonal
ditches, Compared with O. rufipogon^ it is more gregarious and
frugiferous, flowers more synchronously, and has bolder spikelets and
kernels. Early man settled and developed agriculture in drier regions (not
in swamps). It is therefore highly probable that early man relied upon O.
nivara (not O. rufipogon) for developing a grain crop. O. nivara is
"harvested" in large quantities even today by tribal and economically
backward people of central India for self-consumption as well as for
marketing at a premium price as religious people prefer to consume this
"God-given rice" (deohhat) on days of fasting.
The nivara hypothesis proposed by Sharma (1964) and Shastry and
Sharma (1973) and elaborated by Chang (1976) was, however, too simple
to account for all the morphological, ecological, and physiological
variations available in O. sativa. Furthermore, it does not account for the
ecogenetic differentiation in O. sativa and its interecotypic sterility.

S.D. Sharma et al. 357
RECENT STUDIES
Biswal (1988) crossed various collections of O. nivara of India among
themselves and observed increasing Pj pollen sterility with increasing
spatial separation of nivara populations. She/ therefore, concluded that
the pollen sterility observed in the Fi hybrids between japónica and
indica was already present in the progenitor species (O. nivara) and has
merely been carried forward to the progeny species (O. sativa). When
collections of O. nivara were crossed with O. rufipogon, the hybrids
were more fertile than many of the nivara x nivara hybrids. When O.
nivara was crossed with various ecotypes of O. sativa, she observed that
(with regard to pollen fertility) aus x nivara and japónica x nivara hybrids
behaved more or less like nivara x nivara hybrids. The aman x nivara and
javanica x nivara hybrids behaved like rufipogon x nivara hybrids. This
led her to conclude that (a) aus and japónica ecotypes have originated
directly from two different populations of O. nivara, (b) introgression of
rufipogon characters into aus might have given rise to aman ecotype.
Based on the views of Chang (1985), Glaszmann and Arraudeau (1986),
and her own observations, she proposed that migration of hill rices of
mainland Southeast Asia ("hrmsea") to Indonesia followed by
introgression of rufipogon genes into it, could have given rise to javanica
types.
Biswal (1988) assumed that early man worked on different
populations of O. nivara at different sites in southeast India, southwest
China and Southeast Asia for domestication of rice. In other words, plural
sites of domestication from different populations of O. nivara is more
probable as suggested by Harlan (1975) in his h)rpothesis of diffused
origin of agriculture.
Second (1982) analyzed 40-isoenzyme loci of 468 collections of O.
sativa obtained from many countries. On the basis of their pollen
sterility, he could identify two small groups of varieties, which he called
"ancestral" japónica and "ancestral" indica. Assuming the electromorphs
of these two "ancestral" groups to be "parental", he presumed that the
electromorphs of other varieties could be hybrid polymorphs. The
electromorphic diversity of the wild rices is greater than that of the
cultivated rice (Shahi et ah, 1969; Pai et aL, 1973, 1975; Second and
Trouslot, 1980). He concluded that among the various phylogenetic
relationships of rice varieties put forward in the literature, only the
hypothesis of the independent domestication of indica and japónica types
proposed by Chou (1948) fits the observed pattern of isozyme variation.
He further concluded that the pollen sterility between the indica and
japónica subspecies could have existed before domestication.
Tripathy (1994) made many "seihr" x "seihr" crosses and observed
that their F^ hybrids show a wide range of pollen sterility, as observed

358 Rice Breeding and Genetics: Research Priorities and Challenges
by Biswal (1988) in nivara x nivara hybrids. Similarly^ the "seihr" x
japónica Fj hybrids showed maximum pollen sterility (14-94%) among
the various interecotypic combinations and in this sense behaved similar
to nivara x japónica hybrids of Biswal (1988) and japónica x indica
hybrids reported by others (see Oka, 1964; Sampath, 1964; Shastry, 1964
for review). It was therefore evident that the southeast Indian hill rice
("seihr") behaved like O. nivara in its intraecotypic as well as
interecotypic hybrids. She, thus concluded that these southeast Indian
hill rices of Jeypore tract and its adjacent areas might have directly
originated from 0. nivara.
Tripathy (1994) crossed "seihr", japónica, aman, shall and tjereh
cultivars in all inter-ecotypic combinations. Based on the morphology
and pollen sterility of their Fj hybrids, it was observed; that (a) the
characters of japónica ecotype showed dominance in all its interecotypic
hybrid combinations (except in aman x japónica) and (b) the characters
of "seihr" showed dominance over those of all the other ecot3rpes in all
its inter-ecotypic combinations (except "seihr" x japónica). It was,
concluded that the japónica and "seihr" have retained more dominant
genes than the other ecotypes and probably represent two primary
ecotypes from which the other ecotypes have evolved. Within these two
basic ecotypes, the characters of japónica showed dominance over that
of "seihr" and hence the former may represent a more primitive ecotype
than the latter.
The shall ecotype of the Brahmaputra valley is similar to the aman
ecotype with regard to its ecological (and agronomic) preferences and
photosensitivity. The shall types show medium fertility with japónica
and very low fertility with aman and tjereh types. The sterility of aman x
shall and shall x tjereh hybrids vis-a-vis better fertility of shall x japónica
hybrids indicate that the japónica-like forms available in the sub-
Himalayan belt of the Brahmaputra valley might have played a role in the
origin of this ecotype (Tripathy, 1994). The pollen fertility of aman x
tjereh hybrids was the highest among all the interecotypic hybrids of
tjereh. This indicates that the genetic origin of both these ecotypes could
be the same (Tripathy, 1994).
PROPOSED HYPOTHESIS
O. nivara occurs frequently and abundantly in the northeastern part of
the Deccan peninsula (which also includes the Jeypore tract of Orissa)
and in the central Gangetic plains. Its occurrence in western India is
neither so frequent nor so abundant. It is sparsely distributed in southern
India and is conspicuously absent in Gangetic Bengal as well as in the
whole of northeastern India. O, nivara is available again in the plateau

S.D. Sharma et al. 359
regions of Myanmar, Thailand, Cambodia, and Laos and in the
southwestern parts of mainland China. The geographic distribution of
O. nivara, is thus disjunct; one found in India and the other in Indochina
and China.
The small ditches and seasonal pools in the plateau regions of South
and Southeast Asia and southwestern China provide ideal habitats for
O. nivara. Its distribution in these regions must have been more frequent
and abundant when human population was very limited ¿nd hence
cultivated fields were fewer. The hilly tracts were ideal areas for
habitation by Neolithic hunting-gathering man who domesticated many
plants including rice. Morishima (1984) rightly points out that "the deltas
of big rivers were not accessible for early people. Apparently, the hilly
area seems to have played an important role in making contact with rice".
It is therefore highly probable that the people of the northeastern Deccan
plateau domesticated the "seihr" types and the people of southwestern
China japonica-like types from populations of O. nivara of their
respective regions.
Origin of Basic Types
Many tribes belonging to Proto-Australoid ethnic stock inhabit the
Jeypore tract of Orissa, India. These people have been "harvesting" O.
nivara, which occurs naturally and frequently in seasonal ditches, for
ages. They also practice shifting cultivation and grow many primitive
cultivars of rice. With increase in population and dwindling forest
cover, however, they are giving up shifting cultivation and adapting
upland rice cultivation but continue to patronize their age-old rice
cultivars. Ramiah (1953) was impressed with the varietal diversity of
this area and proposed that it probably represented another independent
center of origin of cultivated rice. During 1955-60, the Central Rice
Research Institute, Cuttack collected more than 1,700 traditional
cultivars of rice from this area. 'Oka and Chang (1962) studied these
cultivars of the Jeypore tract of Orissa and regarded them as forms
intermediate between cultivated and wild types "still staying in the
midst of differentiation".
The primitive upland rice cultivars of the Jeypore tract have many
special features such as short height, thin culm, few tillers, small
panicles, and often (though not always) black husk, red kernel, and awn.
They are short-duration, photoperiod-insensitive cultivars. In fact,
similar types of rice cultivars are often cultivated as a rainfed upland
crop especially in unbunded fields in Chhattisgarh, western Orissa and
southern Bihar (Jharkhand) by resource poor farmers. These land races
(locally known as tikradhan, bhatadhan, garodhan, etc.) have been
collectively referred to as "southeast Indian hill rices" or by their

360 Rice Breeding and Genetics: Research Priorities and Challenges
acronym "seihr" in this paper. As already, discussed these ^'seihr" types
have retained many primitive features, express dominance for many of
their characters in their hybrids with other ecotypes and, with regard
to sterility of these hybrids, behave like O. nivara. The ''seihr" types may,
thus, be assumed to have originated directly from O. nivara of southeast
India.
According to Sharma ef al. (1971), Hakim and Sharma (1974),
Asthana and Majumdar (1981) and Sharma (1982), there are various
grades of japonica-like cultivars that are cultivated by the tribal people of
northeast India particularly of Arunachal Pradesh. The higher in altitude
one moves up in the Himalayas, the greater one finds the expression of
japónica traits in its rice cultivars. The landraces in Tawang district of
Arunachal Pradesh (India) probably present a unique example of
japonica-like rice cultivars that are grown at lower latitude (27"N) but
high altitude (1,800 m). The situation in south and southwest China is in
no way different. One can find in the hills of Yunnan and Kweichow
provinces (China), where the sinica ("keng") rices were grown at
elevations above 1,800 m, "a mixture of sinica and indica rices growing
at medium elevations and exclusively indica ("hsien") rices at altitudes
below 1000 m" (Ting, 1961 quoted by Chang, 1985). The japonica-like
types which were domesticated in southwest China spread westward
along the sub-Himalayan belt up to Nepal (or even beyond) and
southward up to Myanmar and Indo-China. According to Watabe et ah
(1970) these jap'onlca-like types had a much more southerly distribution
in Indochina in the first millennium A.D. Within the area of distribution
of japonica-like types, O. nivara occurs frequently in southwest China
(Shao et al, 1986). It is therefore highly probable that the japonica-like
cultivars were domesticated in southwest China.
The hill rices of mainland Southeast Asia ("hrmsea") as described by
Chang (1986) are morphologically similar to bulu and gundil types of
Indonesia though the Indonesian types are late in maturity, have a long
vegetative phase, and are adapted to irrigated agriculture. Genetically,
the japónica and javanica are closer to each other and produce fairly
fertile hybrids when intercrossed (Terao and Mizushima, 1944; Oka,
1958). Based on isoenzyme studies, Glaszmann (1985) put the japónica,
"hrrnsea", and the javanica in the same cluster. According to Glaszmann
and Arraudeau (1986), the morphological characters of the rice cultivars
of Japan, Korea, and northern China and that of Indonesia form two
extremes of the same geographical dine and the cultivars of the montane
areas of Southeast Asia and the Himalayas occupy an intermediate
position. It is therefore probable that the hill rices of mainland Southeast
Asia (''hrmsea") have originated from O, nivara of that region. If so, the
hill regions of the mainland Southeast Asia represent another center (or a

S.D. Sharma et a l 361
subcenter?) of geyietic diversity of 0 . nivara as well as of origin of
cultivated rice, particularly of the "hrmsea" types which seem to be the
progenitors of the javanica ecotype. If O. nivara of southern China and
Southeast Asia are assumed to be genetically closer to each other when
compared with that of Southeast Asia, the closer relationship among the
japónica, “hrmsea", and javanica can be easily explained.
As already discussed, the populations of O, nivara are often
genetically differentiated (Biswal, 1988). This differentiation increases
with increase in spatial separation of their populations and is expressed
as sterility of their hybrids. The genetic differentiation that existed in
the original populations of O. nivara of southeast India and southwest
China has been carried over to the domesticated rice (O. sativa) and is
reflected as sterility in their interecotypic hybrids of O, sativa, e.g. in
japónica x índica hybrids. The "seihr" tjpes of southeast India, the
japonica“like types of southwest China and the "hrmsea" types of central
Indochina may, thus be assumed to represent three basic stocks of O.
sativa that have evolved directly from the annual wild species (O. nivara)
of their respective regions in Asia.
Origin of Primary Ecotypes
The photoperiod-insensitive rice cultivars that are grown in bunded
fields during the monsoon (July-October) season and mature in 100 to
120 days are collectively known as aus types in Bengal. In fact, cultivars
similar to aus are widely cultivated in the whole of southeastern,
northeastern and eastern India although they are called by different
names in different states. Ecogenetically, they are one and the same
group that has been termed as aus in rice literature. The aus cultivars are
genetically superior to "seihr" types in their yield attributes and respond
better to agronomic practices. The aus ecotype seems to have evolved
directly from the upland rice ("seihr") of southeast India. Traditionally,
aus were grown only under rainfed conditions as the whole of eastern
India used to receive sufficient rain. Subsequently, its cultivation spread
from southeast India to other parts of India,
The japonica-like types were carried from southwestern China to
eastern and then to northern China, where they developed into what is
now known as keng types. These keng types had better yield attributes,
were better amenable to agronomic manipulations, and suited to irrigated
conditions.
m
Origin of Secondary Ecotypes
The migration of early man from uplands toward lowlands of river
deltas must have been a later event in the history of rice cultivation

p
362 Rice Breeding and Genetics: Research Priorities and Challenges
(Whyte^ 1972; Morishima, 1984). If so, the primitive cultiváis of rice
evolved from O. nivara were carried by man to new habitats closer to
that of 0 . rufipogon resulting in the introgression of rufipogon genes into
sativa cultiváis.
The aman types of rice cultiváis could be the result of introgression
of rufipogon genes into sativa cultiváis. The aman types are photoperiod
sensitive, late maturing, adapted to wetland rice cultivation, and more
productive than aus types. The lower Gangetic valley was probably the
meeting ground, where genes of O. rufipogon got introgressed into aus
and as a result, aman types were developed. Endowed with the new
traits, the rice plant was capable of spreading over to wetland, an
ecosystem hardly ever exploited by any crop plant.
The shall types, though morphologically as well as ecologically
exhibiting many similarities with the aman types, do show affinity with
the japónica types for some characters such as grain type, number of
priinary branches per panicle, tolerance to cold, etc. In the interecotypic
hybrids reported here, the characters of japónica expressed dominance
over that of "seihr" and the characters of shall over that of aman. In
japónica x shall hybrids also the characters of japónica were dominant.
With regard to pollen fertility of interecotypic hybrids, the shali types
showed greater fertility with japónica (60.73%) than with aman (51.09%).
It is therefore interpreted that shali types are the products of introgression
of rufipogon characters into japonica-like forms of that region. In this
context, it is noteworthy that O, rufipogon and shali cultiváis of O. sativa
are sympatric in the Brahmaputra valley and hence the introgression of
rufipogon genes into the background of japonica-Uke cultiváis is highly
probable.
The migration of hill rices of mainland Southeast Asia ('^hrmsea"
from that region to Indonesia with introgression of some genes of O.
rufipogon ?) could be the only plausible explanation for the physiological
and ecological adaptation of javanica types to irrigated as well as the high
fertility observed in javanica x nivara hybrids by Biswal (1988).
According to Watabe, the Indians carried aman types to Indochina
sometime in the 9th century AD. The tjereh types of Indonesia could
have developed from the aman types of India carried to Indonesia by the
Indians during this period. This may explain the high pollen fertility of
aman x tjereh hybrids reported by Morinaga (1968) and Tripathy (1994).
The primary ecot)^es of O. sativa have retained the photoinsensitivity
of the annual wild species (O. nivara) and man has
successfully exploited this trait to develop genotypes suitable for
cultivation of rice in different seasons of the year. The secondary ecot)rpes
have acquired photoperiod sensitivity and adaptation to lowland and
even greater depths of water. Man has successfully exploited these

L i -
S.D. Sharma et al. 363
secondary traits of the rice plant to spread its cultivation to new
ecosystems.
SUPPORTING EVIDENCE
Our hypothesis that the Asian cultivated rice (O. sativa) had not only a
polyphyletic origin (two species of Oryza have played key roles in its
origin)^ but also a polytopic origin (rice originated independently at
plural sites), does not contradict the findings of other disciplines.
Archeological excavations indicate the presence of rice in the food
economy of early man as far back as 5000 BC in China as well as in India.
This suggests that rice might have been simultaneously domesticated at
many sites. Anthropologically, the whole area starting from the western
coast of India to the eastern coast of southern China and Vietnam was
inhabited during this period by Pro to-Australoids, i.e., by people
speaking the Austric group of languages. Pro to-Australoids practiced
primitive methods of rice cultivation and must have been responsible
for the origin of "seihr types in southeast India, japonica-like types in
Southwestern China and "hrmsea" types in Southeast Asia. However,
rice seemed to have played only a marginal role in their food economy
and was not a staple diet (Whyte, 1972; Kumar, 1988).
Rice could have become a staple diet only after development of the
aman ecotype, which was adapted to a lowland ecosystem and had
greater productivity. For large-scale cultivation of this ecotype, iron
implements (for plowing the land) and draft animals (oxen in India,
water buffaloes in China) must have been prerequisites. This could have
been possible only after the movement of Aryans into the lower Gangetic
valley and the Chinese civilization to southern China. Development of
aman and shall types in the valleys of the Ganga and the Brahmaputra
respectively, must thus have been quite late in the history of
domestication of rice. The aman rices were carried to Indonesia and
Indochina by the Indian coloiuzers around the 9th century AD. The tjereh
types of Indonesia are probably modified forms of aman rices carried
from India to Indonesia.
Watabe and his associates ( Watabe and Akihama, 1968; Watabe,
1970,1973; Akihama and Watabe, 1970; Watabe, ei'ah, 1970; Watabe and
Toshimitsu, 1974; Watabe, et ah, 1976) have extensively surveyed the
rice grains found in the ancient bricks at historical sites of India,
Myanmar, Thailand, Laos, Cambodia and Vietnam. Summarizing their
findings, Watabe et al., (1976) noted that in Indochina there were two
routes of dispersal of cultivated rice in early times, one followed the
Mekong River from Laos to the South, The strains of rice transmitted

364 Rice Breeding and Genetics: Research Priorities and Challenges
along this route showed the characteristic features of the japónica or
japonica-Uke grain type. This type of rice was considered to have been
first cultivated in Indochina. The second route was from India over the
Bay of Bengal to the coastal areas of Indochina. The grain type
transmitted by this route was clearly of the indica or aman type. This
group was transmitted to Indochina at a later date than the japónica or
japonica-like group. The author named the former group ''the Mekong
descent group" and the latter group "the Bengal descent group".
The development of three basic ecotypes of O. sativa from three
different populations of O. nivara in three different regions is not only
associated with their genetic differentiation; but also with their ecological
and physiological differentiation. Tsunoda (1984) inferred the japónica to
bean ecospecies, established in the sub-tropical hardleaf evergreen forest
region grown under a watered condition to avoid the cold, the javanica
established in the tropical rain forest region primarily under rainfed
upland conditions benefiting from the warm climate and the rainfall
throughout the year^ and the indica an ecospecies established in the
monsoon moist deciduous forest region grown under high temperatures
to summer monsoon rains on the uplands and monsoon rains and flood
waters in lowlands forming ecotypes such as early aus and late aman.
References
Akihama, T. and Watabe^ T. 1970. Geographical distribution and ecotypic differentiation of
wild rice in Thailand. T o n a n A j i a K e n k y u ( T h e S o u th e a s t A s i a n S t u d i e s ) B ( 3 ) : 337-346.
Andrews, F,W, 1956. The flowering plants of the Sudan. Vol. III. T. Bunde & Co. Ltd.,
Arbroath, Scotland.
Asthana, A. N. and Majumdar, N. D. 1981. Studies in the rice germplasm of northeastern hill
region. Res. B u l l N o . II, ICAR Complex for NEH Region, Shillong.
Biswal, J. 1988. Genetic differentiation in O r y z a n i v a r a . Unpubli. Ph.D. thesis, Utkal
University, Bhubaneswar, Orissa (India).
Bor, N.L. 1960. T h e G r a s s e s o f B u r m a , C e y l o n , I n d i a a n d P a k i s t a n (excluding Bambuseae).
Pergamon Press, Oxford.
Chang, T. T. 1976. The origin, evolution, cultivation, dissemination and diversification of
Asian and African rices. E u p h y t i c a 2 5 :425-441.
Chang, T. T. 1985. Crop history and genetic conservation : Rice— case study. I o w a S t a te } .
Res. 59(4) : 4 2 5 -A 5 6 .
Chang, T.T. ei a l., 1986. Genetic studies on the components of drought resistance in rice
{ O r y z a s a t iv a L.). In: Rice G e n e t ic s . IRRI, Manila, Philippines, pp. 387-398.
Chatterjee, D. 1948. A modified key and enumeration of the species of O r y z a l A i m . I n d i a n /.
A g r i c , S e i. 18 :185-192.
Chatterjee, D. 1951. Note on the origin and distribution of wild and cultivated rices. I n d i a n J.
G e n e t . P I . a n d B r e e d . 2:18-22,
Chevalier, A. 1932. Nouvelle contribution A l'étude systématique des O r y z a . R e v . B o t . A p p l
d ' A g r i c . T r o p . 12 :1014-1032.

S.D. Sharma e t a l , 365
Chou, C.L. 1948. China is the place of origin of rice. /■R i c e S o c . C h i n a 7 : 53-54 (in Chinese).
Clayton, W. D. 1968. Studies in the Gramineae. XVII. K e w B u l l . 21(3): 485-488.
De Candolle 1886. O r i g i n o f C u l t i v a t e d P l a n t s , (fase, ed.). Haiher, New York, 1967.
Dhua, S. R. 1994. Genome analysis of O r y z a r h iz o m a t i s Vaughn. Ph.D. thesis, Visva Bharati,
Sriniketan (India).
Glaszmann, J.C, 1985. A new insight into Asian cultivated rice classification from isozyme
studies. Rice Genet. Neioslett. 2:48-51
Glaszmann, J. C. 1986. A varietal classification of Asian cultivated rice {Oryza sativa L.)
based on isozyme polymorphism. In: Rice Genetics. IRRI, Manila, Philippines, pp, 83-90.
Glaszmann, J.C. and Arraudeau, M. 1986. Varietal diversity and reproductive barriers, I,
Rice plant type variation, ;aponiCß-/flWtticß relationships. RiceGanet. Newsktt. 3 :41-43.
Hakim, K. L. and Sharma, S. D. 1974. Localised distribution of certain characters of rice
cultivars in northeast India, Proc. 2nd Gen. Congr., SABRAO. Indian }. Genet. 34A: 16-21
Hamada, H. 1949. Über den Ursprung des Reisbaues in Asien ( in Japanese with German
Summary) P r o c . C r o p S e i. S o c . j a p a n . 18:106-107.
Henderson, M. T. 1964. Cytogenetic studies at the Lousiana Agricultural Experiment Station
of species relationships in O r y z a . In: R ic e G e n e t ic s a n d C y t o g e n e t ic s . Elsevier, A m s t e r d a m ,
pp 103-110.
Harlan, J. R. 1975, C r o p s a n d M a n . S o c i e t y o f A g r o n o m y , Madison, Wisconsin.
Hooker, J.D. 1897. F l o r a o f B r i t i s h I n d i a , Vol. VIII. Reeve & Co, Ltd., London.
IRRI, 1964a. Report of the Committee appointed to attempt a standard classification and
nomenclature of the genus O r y z a . In: R i c e G e n e t ic a n d C y t o g e n e t ic s . Elsevier, Amsterdam,
pp. 251-252.
IRRI, 1964b. Recommendation of the Committee on Genome Symbols for O r y z a Species. In:
R i c e G e n e t ic s a n d C y t o g e n e t ic s . Elsevier, Amsterdam, pp. 253-254.
Kumar, T. T, 1988, H i s t o r y o f R i c e i n In d i a ; M y t h o l o g y , C u l t u r e a n d A g r i c u l t u r e . Gian Publ.
House, Delhi.
Linnaeus, C. 1753. Species Plantarum, vol. I. Holmaiae, Laurentii, Salivi.
Morinaga, T. 1968. Origin and geographical distribution of Japanese rice. J a p . A g r i e . R e s ,
Qtrly. 3 : 1-524.
Morishima, H. 1984. Wild plant and domestication. In: B i o l o g y o f R i c e . S . Tsunoda and N.
Takahashi (eds,). Elsevier, Amsterdam, pp. 3-30.
Oka, H. 1,1958, Intervarietal variation and classification of cultivated rice. I n d i a n J . G e n e t 18
: 79-89.
Oka, H. 1 .1964. Pattern of interspecific relationships and evolutionary dynamics in O r y z a .
s a tiv a L. and five wild diploid forms o í O r y z a . C r o p S e t. 1:445-450.
Oka, H. 1.1974. Experimental studies on the origin of cultivated rice. G e n e t ic s 78:475-486.
Oka, H, 1.1988. O r i g i n o f C u l t i v a t e d R ic e . Elsevier, Amsterdam.Jap. Sei. Soc., Tokyo.
Oka, H. I. and Chang, W. T. 1959. The impact of cultivation on populations of wild rice,
O r y z a s a t iv a f. s p o n ta n e a . P h y t o n . , 13:115-117.
Oka, H. I. and Chang, W. T, 1962. Rice varieties intermediate between wild and cultivated
forms and the origin of t h e ja p ó n i c a ty p e . B o t . B u l l . A c a d . S ín ic a 3 : 109-131.
Pai, C., Endo T. and Oka, H. 1.1973. Genetic analysis for peroxidase isozymes and their organ
specificity in O r y z a p e r e n n i s and O. s a t iv a . C a n . /. G e n e t C y t o t . 15:845-853.
Pai, C., Endo, T. and Oka, H. 1.1975. Genetic analysis for acid phosphatase isozymes in O.
p e r e n n is a n d O , s a t iv a . C a n . J . G e n e t a n d C y t o l . 1 7 :637-650.
Porteres, R. 1956. Taxonomie agrobotanique des riz cultives O r y z a . 0. s a tiv a L. et O .
g la b e r r im a S t e u d . J . A g r . T r a p . B o t . A p p l . 3:341-384,541-580,627-700,821-856.
Prain, D. 1903. B e n g a l P l a n t s , Vol.IL N.W. & Co., Calcutta.

366 Rice Breeding and Genetics: Research Priorities and Challenges
Ramanujam, S. 1938. Cytogenetic studies in the Oiyzeae, III. Cytogenetical behaviour of an
interspecific hybrid in Oryza. J. Genet. 35 :223-258.
Ramiah, K. 1953. Rice Breeding and Genetics. Sci. Monog. No.l9, Indian Council Agrie. Res.,
New Delhi.
Ramiah, K. and Chose, R. L .M. 1951. Origin and distribution of cultivated plants of south
Asia ; Rice. Indian J. Genet. 11 ; 7-13.
Richharia, R. H. 1960. Origin of cultivated rices. Indian J. Genet. 20:1-14.
Roschevicz, R. J. 1931. A contribution to the knowledge of rice. Appl. Bot. Genet. PI. Breed.
Bull. 27:2-133 {in Russian with English summary).
Roxburgh, 1832. Flora Indica, Vol.II, 200
Sampath, S. 1962. The genus Oryza ; its taxonomy and species inter-relationships. Oryza 1:1-
29.
Sampath, S. 1964a. The significance of hybrid sterility in rice. In: Rice Genetics and
Cytogenetics. Elsevier, Amsterdam, pp. 175-186.
Sampath, S. 1964b. The species ancestral to cultivated rice. Curr. Sci. 33(7) : 205-207.
Sampath, S. and Rao. M. B.V. N. 1951. Inter-relationship between species in the genus Oryza.
Indian J. Genet. 11:14-17.
Sampath, S. and Govindaswami, S. 1958. Wild rices of Orissa, their relationship to cultivated
rices. Rice News Teller. 6 : 17-20.
Second, G. 1982. Origin of the genic diversity of cultivated rice (Oryza spp.): Study of the
polymorphism stored at 40 isozyme loci. }pn. /. Genet. 57:25-57.
Second, G. and Trouslot, P. 1980. Polymorphisme de trieze zymogrammes observes parmi
diverses especes sauvages et cultivées du genre Oryza. In: Electrophorèse d'enzymes de riz
(Oryza spp.). Travaux et documents 120: 50-58, ORSTOM, Paris.
Senara tna, J. E. 1956. The Grasses of Ceylon. Govt. Press, Colombo.
Shahi, B.B., Morishima, H. and Oka, H.I. 1969. A survey of variations in peroxidase acid
phosphatase and esterase isozymes of wild and cultivated Oryza species. Jpn. J. Genet.
44 :303^19.
Shao, Q., Yi, H. and Chen, Z. 1986. New finding concerning the origin of rice. In: Rice Genetics.
IRRI, Manila, pp. 53-58.
Sharma, S. D. 1964. Interspecific Relationship in Genus Oryza. Unpubl. Ph.D. thesis, Indian
Agrie, Res. Inst, New Delhi.
Sharma, S, D. 1982. Collection and evaluation of rice germplasm from northeast India.
IBPGR Plant Genet, Resources Newslett. 50 ; 62-69.
Sharma, S. D. 1986. Evolutionary trends in genus Oryza. In: Rice Genetics. IRRI, Manila,
Philippines., pp. 59-67.
Sharma, S. D. and Shastry, S. V. S. 1965a. Taxonomic studies in genus Oryza, I. Asiatic types
of O, sativa complex. Indian }. Genet. 25 : 145-156.
Sharma, S.D. and Shastry, S.V.S. 1965b. Taxonomic studies in genus Oryza, II. O. rufipogon
Griff, sensu stricto and O. nivara Sharma et Shastry nom. nov. Indian }. Genet, 25:157-165.
Sharma, S. D. and Shastry, S.V.S. 1966a. Taxonomic studies in genus Oryza, III. Some
nomenclatural confusions. BmÍ/. Bot. Surv. India 6:211-218.
Sharma, S. D. and Shastry, S.V.S. 1966b. Taxonomic studies in genus Oryza, V. Sderophyllum
coarctatum (Roxb.) Griff, Bull. Bot Surv. India 8:42-44.
Sharma S. D., Vellanki, J.M.R., Hakim,, K.L. and Singh, R.K. 1971. Primitive and current
cultivars-of rice in Assam—a rich source of valuable genes. Cwrr, Sci. 40:126-128.
Shastry, S.V.S. 1964. Is sterility genic in japonica-indica rice hybrids In: Rice Genetics and
Cyio^enefics. Eisevier Amsterdam, pp. 154-157.

S.D. Sharma et ah 367
Shastry^ S. V. S; and Sharma, S. D. 1973. In; Evolutionary Studies in World Crops: Diversity and
Change in the Indian Sw^conimeni. J. Hutchinson (ed.). Cambridge Univ. Pres, London, pp.
55-63.
Shastry, S.V.S., Sharma, S.D. and Ranga Rao, D. R. 1961. Pachytene analysis of Oryza, III.
Meiosis in an intersectional hybrid. Nucleus 4 : 67-80,
Takahashi, N. 1984. Differentiation of ecotypes inOryzfl sativa L. In: Biology of Rice S. Tsunoda
and N. Takahashi (eds.). Elsevier, Amsterdam, pp. 31-67.
Tateoka, T. 1962. Taxonomic studies of Oryza, II. Several species complexes. Bpt. Mag, 75 :
455-461.
Tateoka, T. 1965. Porteresia, a new genus of Gramineae. Bull Natl. Sei. Mus. (Toicyo) 8: 405-
406.
Terao, H. and Mizushima, U. 1944. On the affinity of rice varieties cultivated in east Asia
and America. Bull Agric. Expt. Sta., Ministry of -Agric. and Commerce (Japan) 55:1-7 (in
Japanese).
Ting, Y.(ed.). 1961. Cftinese Culture of Lowland Rice (in Chinese). Agric. Publ. Soc., Beijing.
Tripathy, S. 1994, Ecogenetic Differentiation in Oryza sativa L. Unpubl. Ph. D. thesis, Utkal
Univ,, Bhubaneswar, Orissa (India).
Tsunoda, S. 1984, Synthesis and perspectives. In: Biology of Rice. S. Tsunoda and N. Takahashi
(eds.). Elsevier, Amsterdam, pp. 361-375.
Vaughan, D. A. 1994, The Wild Relatives of Rice. IRRI, Manila, Philippines.
Watabe, T. 1970. The alteration of cultivated rice in Thailand and Cambodia. Tofiim Ajia
Kenkyu (The Southeast Asian Studies) 8 :36-45.
Watabe, T. 1973. Alternation of cultivated rice in Indochina. Jap. Agric. Res, Qtrly, 7(3): 160-
163.
Watabe, T. and Akihama, T. 1968. Morphology of rice grains recovered from ruins in
Thailand. Tonan Ajia Kenkyu (The Southeast Asian Studies) 6(2): 331-334.
Watabe, T. and Toshimitsu, K. 1974. Morphological properties of old rice grains recovered
from ruins in Indian sub-continent, A study on the alterations of cultivated rice.
Preliminary Report of Tottori University's Scientific Survey, Voi. 2, pp. 1-18.
Watabe, T., Akihama, T. and Kinoshita, 0 . 1970. The alteration of cultivated rice in Thailand
and Cambodia. Tonan Ajia Kenkyu {The Southeast Asian Studies) 8(1): 36-45
Watabe, T., Tanaka, K. and Nyunt, K, 1976. Ancient rice grains recovered from ruins in
Burma. A study on the alteration of cultivated rice. Preliminary Report of the Kyoto
University Scientific Survey to Burma, pp. 1-18.
Watt, G. 1891. Dictionary of the Economic Products of India, Vol. 5, pp. 498-654.
Whyte, R .0 .1972. The Gramineae, wild and cultivated in monsoonal and equatorial Asia, I.
Southeast Asia. Ancient Perspectives V. 15(2): 127-151.
Yeh, B. and Henderson, M. T, 1961. Cytogenetic relationship between cultivated rice, Oryza
sativa L. and five wild diploid forms of Oryza. Crop. Scl 1; 445-450.
APPENDIX
Nomenclature of Some Oryza Species
1. O. harthii A. Cheval This is the annual wild species of Africa with a
genomic constitution of AA. During the 1950s and 1960s, this species
Sharma, S.D., Tripathy, S. and Biswal, J. 1997. Origin of Asian cultivated rice and its
ecotypic differentiation. Indian }, Genet. & Plant Breed. 57(4): 339-360.
The authors disown the paper due to extensive textual, grammatical and typographical
distortions made by the editor of the journal.

m
368 Rice Breeding and Genetics: Research Priorities and Challenges
was known to rice researchers by its synonym O. breviligulata A. Cheval.
et Roehr. The binomial O. barthii was wrongly used for the perermial
wild species (0. longistaminata) by almost all biosystematists until
Clayton (1968) clarified the situation.
2. O. glaberrima Steud. The African cultivated species of rice. Its
cultivation has remained confined to tropical west Africa.
Morphologically, it closely resembles the Asian cultivated species (O.
sativa). The two cultivated species, O, glaberrima and O. sativa, can be
differentiated by a few distinctive characters. The variation available in
the two cultivated species is strikingly parallel. The hybrid between the
two cultivated species is, however, highly sterile. According to Porteres
(1956), p. glaberrima originated from the annual wild species (our O.
breviligulata) in tropical west Africa and is still grown only in that region.
3. O. glumaepetula Steud. The American perennial species with a
genomic constitution of AA. It is widely distributed in tropical America
from Cuba to Paraguay. It is also known as O. cuhensis Ekman, a nomen
nudum, and as O. perennis subsp. cubensis. Some taxonomists (Tateoka,
1962) merge this taxon with the Asian perennial species, O. rufipogon.
The Asian species (O. rufipogon) is a runner, whereas the American
species (O. glumaepetula) is semierect. Subgenomically, they are different
(Henderson, 1964).
4. O. longistaminata A. Chaval. et Roehr. ThisTs the perennial wild
species of Africa with a genomic constitution of AA. It was wrongly
identified as O. barthii by all rice biosystematists until Clayton (1968)
pointed out the mistake. It was referred to as O. perennis subsp. barthii by
those who treated the perennial > wild species {glumaepetula,
longistaminata, and rufipogon) of America, Africa, and Asia as a single
species.
5. O. minuta J.S. Presl. ex C.B. Presl. This is a tetraploid {2n = 48) wild
species with a genomic constitution of BBCC, Its distribution is restricted
to the Philippines only. Morphologically, it resembles the diploid {In ~
24) species O. officinalis (genomic constitution CC or DD?) which is
widely distributed in South and Southeast Asia. Taxonomists, earlier
unaware of this cytogenetic difference, merged the two species and
called the combined species O: minuta as this binomial has priority over
the binomial O. officinalis.
6. O. nivara Sharma et Shastry. This is the annual wild species of Asia
with a genomic constitution of AA. It is closely related to the Asian
cultivated species, O. sativa. It has also been known as O. fatua Koenig
nomen nudum. It was identified as O. sativa van fatua following Prain
(1903) and as O. sativa var. spontanea following Roschevicz (1931), which
also included their naturally occurring hybrids. Some taxonomists
(Tateoka, 1962) merge this annual species with the perennial wild species

S.D. Sharma et a l 369
O, rufipogon. Senaratna (1956) and Sampath (1962, 1964b) wrongly
identified it as O. rufipogon Griff.
7. O. officinalis Wall, ex Watt. This is a diploid {2n - 24) species with a
genomic constitution of CC. According to Dhua (1994), its genomic
constitution should be DD. It is widely distributed in South and Southeast
Asia. It grows in partial shade in forests near running streams, moist
grounds, and sometimes in shallow ditches. The hybrid between O.
sativa and O, officinalis is completely sterile. Morphologically, it closely
resembles the tetraploid species O. latifolia (genome CCDD) of America
and O, minuta (genome BBCC) of the Philippines. It was therefore
misidentified as O. latifolia (Hooker, 1897) or as 0 . minuta (Bor, 1960).
8. O. rufipogon Griff. This is the diploid perennial wild species of Asia
with a genomic constitution of AA. It is closely related to the annual wild
species (G. nivara) and the cultivated species (O. sativa) of Asia. It was
identified as O. perennis Moench by all rice biosysternatists from 1951 to
1962 following Chatter]ee (1948) until Bor (1960), Tateoka (1962), and
Sharma and Shastry (1965b, 1966a) clarified the position. It has also been
referred to as O. perennis subsp. balunga (Sampath and Govindaswami,
1958) or as O. balunga (Yeh and Henderson, 1961). According to Tateoka
(1962), O. rufipogon (sensu lato) includes O. nivara and O. glumaepetula.
9. Porteresia coarctata (Roxb.) Tateoka. This is a tetraploid (2u = 48)
species that grows in the tidal swamps of rivers of South Asia. It was
known as Oryza coarctata Roxburgh to all rice researchers until Tateoka
(1965) removed it from the genus Oryza and erected a new genus,
Porteresia, to accommodate this single species. Sharma and Shastry
(1966b) have provided additional evidence for its removal from the
genus Oryza. Its genome has not been determined so far.
10. O. eichingeri. This is a small plant growing in forest shades of
equatorial Africa. It is a diploid (2n = 24, genome = CC) species. Tateoka
collected it in 1964 for the first time in the living form. It also occurs in
Sri Lanka. The Sri Lankan form was earlier known as O. officinalis
(Ceylon) and was used for genome analysis, established to be CC. This
Sri Lankan form was later treated as a distinct species and called O.
collina.
11. O. schweinfurthiana Prod. This is a tetraploid species with a
genomic constitution BBCC. It is widely distributed in tropical Africa. It
grows in partial shade near running streams or pools of water.
Roschevicz (1931) and Andrew (1956) treat O, schweinfurthiana Japanese
authors fallow Tateoka (1962). Earlier, if was misidentified as O.
eichingeri until Tateoka (1962) pointed out the mistake (2n = 48) and O.
punctata (2n = 24) as two species. Tateoka (1962), however, merged the
two species and to differentiate designated them as O. punctata {2n = 24)
and O. punctata (2n = 48).

Index
Abiotic 1,14,46, 73,263,272,273,306
Abortion 113
Accessions 334,335,336
Acid lowland 223, 228
Acid sulfate 219, 221,224,230, 231,235
Acid upland 219,222,223,224,228, 230,
232, 235
Acidity 63
Additive 177,205, 233,253
Additive genetic variation 122
Adventitious roots 86,88
Adverse soil 219
Aerenchyma 85,86,87
Aerobic 222, 226, 231, 274, 331
Aerobic soil 289
Agarose gel electrophoresis 334
Agriculture 55,65
Agrobacteriium-mediated
transformation 211
Agrochemical 1, 71
Agroecology 57,67
Agroecosystems 73
Agroforestry 65,69
Alcohol dehydrogenase (ADH) 331
Alien chromosome 307
Alien genes 15,272,274,275,276,277,278,
279,280,282, 283,288
Alien germplasm 273,283
Alien species 234,271
Alkali injury 227
Alkaline soil 223
Alkalinity 221,225,228, 229,230, 231
All India Coordinated Rice Improvement
Program (AICRIP) 158,183,207
Allele 30,110,112,113,210, 243, 245,281,
282
Allelic 148,150,176,195, 200
Allelic interaction 109,110, 111, 114
Allelochemical 153
Allelopathy 1
Alley cropping 66
Alloplasmic 31
Allotetraploid 288, 290,335, 338
Allotriploid 306
Alogamous 30
Aluminum toxicity 224,225,226, 228,
230, 232, 233, 235
Aman 40, 62, 353, 358, 362, 363, 364
Amphidiploids 306, 322, 323, 340
Amplified fragment length
polymorphisms (AFLPs) 242
Anaerobic 86,228
Aneuhaploids 301
Aneuploid 44,279,293
AngustifoUa 313,314,321,339,340
Annual 316,333,339,341,352
Annual cultivated 351,356
Annual wild 351
Annual wild species 350,351,355,356,
361,362,315
Anther culture 44,45,231,301
Anthesis 77
Anthropologically 363
Antibacterial compound 154
Antibiosis 151,152,203, 205
Antixenosis 152
Apomixis 13,34,44,47
Aquaculture 65
Archeological excavations 363
Aus 40,112,361,362,364
Auto triploid 294
Autogamous 30
Autopolyploid 329
Autosyndetic 329
Autotetraploid 276,279,293,306
Auxins 228
Avirulence 171
Avirulent 153
Avoidance 227
Azolla 58

372 Rice Breeding and Genetics: Research Priorities and Challenges
B
Bacillus thuringiensis 15,160
Backcross 36,39,161,175,176, 209
Bacteria 58,59
Bacterial artificial chromosome (BAC) 15
Bacterial leaf blight (BLB) 242, 243,263,
264
Bacterial leaf streak 169
Bacterial sheath rot 169
Bacterial blight(BB) 15,105,107,144,148,
150,151,154,156,157,169,182,183,
184, 185,186,211,244, 271, 273, 274,
275, 278, 279, 289, 290,307
Bactericide 209
Bengal descent group 364
Biodiversity 2, 7, 8, 54, 61
Biolistic method 211
Biological 4,54
Biomass 5,33, 74, 77,100,101, 104, 273,
289
Biometrical 30,131,138
Biosynthesis 87,223
Biosys tema tics 288
Biotechnology 4,14, 7, 32,44, 69,159,161,
209, 233, 234, 236, 241, 271, 272
Biotic 59,225,271,306
Biotic stress 1,273
Biotype 147,150, 153,155,162,194,195,
196,197, 202, 203, 272, 273, 277, 282
Bivalents 292, 314, 327, 328,329
Blast 16,17, 59, 74,105,107,144,146,147,
148.149.150.154.156.157, 161,169,
170,171,172,173,174,175,176,177,
178,180, 211, 229,242, 243, 244, 271,
273, 274, 279,282, 289,290
Blight 59
Blue-green algae 58
Boro 40,62
Boron toxicity 224,229
Brachyanthae 314,321
Brown plant hopper (BPH) 16,17,105,
107.144.145.146.148.149.155.157,
159,160,192,194,196,197,198, 211,
246, 247,271, 273, 274, 275, 277, 278,
279, 281, 282, 289, 306, 307
Brown spot 155,158,169, 186
Bulk method 126,209
Bulu 41,360
Bush fallow 66
Calcareous 219
Calcareous saline sodic 221,231
Carbon assimilation 89
Carbon partitioning 79,80
Cellular genetics 280
Central Rice Research Institute
(CRRI) 151,187,359
Centro International de Agricultura
Tropical (CIAT) 182, 224,232,236
Centromere 292, 293,297,298,299
Chemical control 59
Chitinase 16,17,153,161,189
Chloroplast 77,78,80
ChloroplastDNA (ctDNA) 331,333,
334,335, 336
Chloroplast genome 334,336
Chromosomal interchange 301
Chromosome complement 292
Clone 283,334
Cloning 15,161, 234, 236, 338
CO2 fixation 77
CO2 assimilation 79
Coastal lowland 222
Coastal plains 354
Coastal saline 229
Coastal salinity 220,228,230
Coastal swamp 223
Coat protein (CP) 18
Codominant 210,211
Codon 15
Cold tolerance 14,107
Combining ability 31,33,113,175,233
Complementary dominant genes 190
Complementary gene 150,180,187,189,
202
Complete resistance 181,243
Computer simulation 127
Conservation 55,61,70,83
Consultative Group on International
Agriculture Research (CGIAR) 6,14
Correlation 131,152,153,253
Covariance 138
Crop intensification 62
Crop rotation 67
Cropping intensity 56
Cropping system 62, 66, 70
Cropping systems 5,7,14,104,145
Crystal protein 160
Cultivated rice 354, 361, 315, 316, 332,
333, 337
Cultivated species 315, 316
Cytogenetics 287,288,290,296,301,319,
320,322,332, 334,338

Index 373
Cytoplasm 31, 36, 40,41, 79, 83,148, 275
Cytoplasmic gene 15
Cytoplasmic male sterility (CMS) 13, 34,
36,37, 38, 39, 41, 42, 46, 47,113, 248,
249, 273, 274, 275, 289
D
Dead hearts 204, 205
Deep water 2, 6,14, 25, 64, 65, 70, 73, 87,
88
Deficiency 219, 221, 222, 234
Deforestation 66, 70, 74
Deletions 334,337
Deoxyribose nucleic acid (DNA) 15,265,
280
Diallel analysis 177, 233
Diallel selective mating (DSM) 206
Differentiation 317, 359, 361, 364
Digenic 172,175, 253
Dilatory resistance 146
Diploid 279, 288, 290, 293, 294, 296, 298,
299, 301, 350, 354, 314, 317, 318, 321,
322, 323, 327, 332, 335, 339, 340
Directorate of Rice Research (DRR) 187
Disease 4,13,17,36,42,45,59,69,75,103,
104,105,120,122,135, 143,144,145,
156,158,169,196,197, 209, 210, 234,
246, 271, 272,273
Disease escape 180
Disease forecasting 59
Disease resistance 17,82,102
Disj unction of chromosome 330
Disomic 276, 277, 279,300
Distorted segregation 112
Diversification 63, 67, 71,145
DNA finger printing 161,162
DNA hybridization 290
DNA marker 4,15,280
DNA probe 336
DNA-chip technology 264
Domesticated 360
Domestication 357
Dominance 28, 30,137, 253, 360, 362
Dominance effects 177, 233
Dominant 41,135,148,149,150,151,171,
174,175,176,177,178,179,184,187,
194,195,199,200, 201,202, 210, 233,
275, 282,358
Dominant complementary gene 172
Double cropping 69
Double trisomics 293
Drought 6,13,14, 25, 46, 65, 73, 74, 75, 76,
78, 80, 82, 90, 272, 280, 289
Drought avoidance 82, 83, 273
Drought escape 82
Drought tolerance 82, 83, 273
Duplex genotype 296
Duplication 293
Durable resistance 59
Dwarf 169
Early , generation selection 125,127
Ecogenetic 353, 356, 361
Ecology 54,56,63,70,232,351,325
Ecospecies 353,364
Ecosystem 2, 3,4, 5,6, 7,10, 65, 73, 74, 90;
144,158,181, 362, 363
Ecosystem 65,146
Ecotourism 65
Ecotype 88, 232, 349, 353, 357, 358, 361,
362, 363
Electromorphs 357
Electron acceptors 78
Electron flow 79
Electron transport 78
Electrophoresis 331
Electrophoretic pattern 331
Electroporation 15,16,211
Elongation 14,88,273,289
Embryo 340
Embryo culture 329
Embryo rescue 14,44,45,160,276,277,
279,307
Environment 30,54,55,68
Environmentally sensitive genetic male
sterility (EGMS) 34
Epidemiological 146
Epiphytotic 59,69
Epistasis 30,31,254
Epistatic effects 210
Epistatic interaction 189
Erosion 65
Ethylene 86, 87,88, 89
Etiology 183,187,189,193
Evapotranspiration 80
Evolutionary trends 338
Expressivity 135
Extra chromosome 293,295,296,297,298,
299
False smut 144,145
Farming systems 5, 8, 68, 69, 71
Fauna 59
Female gamete 110

374 Rice Breeding and Genetics: Research Priorities and Challenges
Fertility restoration 35, 248, 249,250
Fertilizers 56,59, 61, 209
Field resistance 146,173,180
Field screening 226
Field tolerance 147
Fishery 69
Floating rice 88
Flood 56,57,87,88
Flood plains 6
Flood water 84,85,89,90
Flooded rice 7,64
Flooding 6,10,74,84
Flood-prone 46
Fluorescence in situ hybridization (FISH)
331, 332
Food and Agriculture Organization
(FAO) 10,11, 54, 55, 71,119, 229, 287
Forestry 55,69
Fossil energy 70
Fossil fuel 3,71
Fungi 58, 59, 243
Fungicides 209
Gall dwarf 201
Gall midge (GM) 17,144,147,148,149,
150.152.155.157.159.162.198, 201,
202,203, 211,246
Gamete 109, 111, 296
Gamete abortion 109, 111, 112
Gametophyte 35,37,41,44, 77
Gene center 184
Gene gun 211
Gene mapping 241
Gene pool 115,161,206,245, 264, 272,
273, 275
Gene tagging 211,241
Gene-for-gene 150,170
Generalized resistance 146
Genetic 1,28,30,31,32,38,42,57,59,60,
99,105,112,113,116,146,155,159,184,
190.198, 200, 201, 202, 205,225, 233,
236,241, 242, 246, 247, 248, 249, 250,
251,252, 253, 254,263,301,361
Genetic barrier 353,316,317,318,319,353
Genetic base 121,125,145,283
Genetic control 58
Genetic correlation 128,129,130,131,
132,137,138
Genetic differentiation 330,364
Genetic distance 108
Genetic diversity 31,46,61,145,197,361
Genetic drift 121
Genetic engineering 160, 211, 265, 283
Genetic erosion 2,60
Genetic imbalance 296, 301
Genetic linkage 296
Genetic male sterile 207
Genetic manipulation 101,241
Genetic map 4,15
Genetic origin 358
Genetic parameters 122
Genetic resistance 120,159
Genetic resource 55
Genetic variability 272
Genetic variance 138
Genetic variation 121,125,132,133,137,
335
Genome 36,160,241,248,254,264,272,
275, 279, 280, 281, 288, 290, 291, 350,
317, 324, 325, 326, 327, 328, 329, 330,
332, 333, 337, 338, 340
Genome evolution 337
Genome mapping 265
Genome analysis 338
Genomic constitution 350,354,355,317,
318, 323, 325, 326,327,335
Genotype 30,35,59,82,86,103,109,110,
114,128,134,135,136,146,170,171,
184,185,210,227, 230,231,232,233,
252, 296,362
Genotype environment interaction 134,
135
Geographic distribution 347,351,359
Geographic origin 335
Geographical race 328
Germplasm 61,82,84,105,107,108,184,
189, 205,206,225, 229,234, 251,252, 273
Gibberellin biosynthesis 87
Gibberellin deficient 88
Gluconase 153
Glycolytic pathway 331
Grain density 103
Grain quality 247,248
Gramineae 287,288
Grandiglumis 190,272,273,289,319,320,
321,323,325,326,327,330,331, 332,
335, 341
Grassy stunt 59,105,169,192,193,273,
274, 275,289
Green house gas 7,12,64
Green leaf hopper (GLH) 105,107,144,
148,149,151,175,189,195,197,199,
200,201,246,273, 274,289

Index 375
Green leaf manuring 66
Green manuring 234
Green revolution 1,2,5,53,54,55,68,
143,145
Gross national product (GNP) 19
Gundhi bug 144
Gundil 360
H
Haploid 14,15,30,301
Haplotype 128
Hardpan 56,75
Harvest Index (HI) 1,2,24,33,100,101,
104
Herbicide 16
Heritability 122,123,138,189
Heritable 132
Heritable variation 123,126
Heterobeltiosis 24,29
Heterochroma tic 292,296
Heterosis 12,23,24,25,28,29,30,31,44,
45, 46, 47,108,110,113,120,122,123,
132,136,137, 251, 252, 253, 254
Heterotic 31,32,33,43
Heterozygote 30,128,129,252
Heterozygous 174
High yielding varieties (HYV) 1,54,56,
60, 61, 62, 63, 68,144,145
Hill rices 362
Hispa 144,152
Hoja blanca 169, 211
Homeostasis 339
Homologous 272,283,295,297,298,301,
306,326,327,329
Homology 327,328,329
Homozygote 30,253
Homozygous 28,128,133,135,174,299
Hopper burn 194,197
Horizontal resistance 146,181
Hormone 89
Host parasite interaction 154,173
Host plant resistance 146, 156,209
Hot spots 182,226,227,231
Hrmsea 353,357,360,361,362,363
Hsien 360
Hybrid 12, 25, 26, 27, 28, 32,33, 34, 37, 39,
42, 43, 44, 47,112,115,120,136,137,
248, 251, 252,264, 275,279,306,318,
323,326,327,323, 326,327,329,357,
358,360, 361,362
Hybrid barrier 116
Hybrid embryo 276
Hybrid polymorphs 357
Hybrid population 355
Hybrid rice 3,12,23,45,60,70,80,108,
110, 249
Hybrid sterility 109,110, 111, 113,114,
115,116,251
Hybrid sterility gene loci (HSGLi) 112,
113,115
Hybrid vigor 23
Hybridization 32,37,116,173,209, 229,
231,296,336, 337,340
Hydrogen sulfide toxicity 224
Hydrophytes 314,339
Hydrophytic 339
Hypersensitive 153
I
Ideotype 102,105,120
Idiogram 291, 298, 299,302
Immune 184
Immunity 146
Inbred 28
Inbreeding 125
Incompatibility 306
Indica 12,32, 33, 37, 38, 41, 42, 43, 46,107,
108, 109,110, 111, 113,115,120,122,
125,127,131,174,175,176,177,229,
232, 243,249,250,251,252, 253,288,
291, 293,294,296,301,331,333,335,
336,337,357,364
Indicoides 341
Inhibitor 17, 82,83,87,88,160,161,172,
187
Inland salinity 220,228,230
Inland swamps 6
Inland valley swamps 6,223,232
Insect pests 4,13,16,42,148,169,193,209
Insect resistance 17,45,102,105, 210,246
Insectide 59
Insects 58,103,105,147,152,155, 234, 271
Insertions 334,337
Integrated nutrient management (INM) 4,
58
Integrated pest management (IPM) 4,58,
59 61,159,162
Intellectual property rights 264
Interallelic 31
Interecotypic 356,358,361, 362
Integenomic 330
Intermolecular 334
International Institute of Tropical
Agriculture (IITA) 232

376 Rice Breeding and Genetics: Research Priorities and Challenges
International Network for Genetic
Evaluation of Rice (INGER) 136, 205,
206, 208, 234
Internarional Rice Acid Lowland
Observation Nursery (IRALON) 234
International Rice Research Institute (IRRI)
10,12,14, 27, 42, 81,100,105,119,120,
121,124,136,148,149,151,181,182,
186, 188,192, 201, 205, 206, 208) 224,
226, 231, 232, 236,. 250, 274, 275, 283, 293
International Rice Salinity and Alkalinity
Tolerance Observation Nursery
(IRSATON) 234
International Rice Stem Borer Nursery
(IRSBN) 207
International Rice Testing Program
(IRTP) 234
Interspecific 36, 44, 275, 276, 277, 278,
279, 306, 358, 360
Interspecific differentiation 336
Interspecific hybrids 324, 325,328, 329,
330
Inter-subspecific 110
Intra-allélic 31
Intraecotypic 358
Intragenomic 330
Intramolecular 334
Introgress 273,274, 275, 307
Introgression 15,121,242,274,275,276,
277, 278, 279, 280, 281,282,283> 288, '
306, 316, 335, 338, 352,357,362
Introgressive hybridization 317,318,353
Iron chlorosis 226,231,235
Iron deficiency 223, 228,230, 231
Iron toxicity 223,225,227,228,229,230,
231, 232, 233,235, 236,263
Irrigated 2, 3,10,11,45,55,56,60, 62, 63,
70, 73,144,181, 223, 360, 361
Irrigation 57,61
Isochromosome 297,298,299
Isoelectric focusing (IBP) 331
Isogenic 31, 39,170, 248
Isogenomic 330
Isolate 243
Isolate specific resistance 151
Isozyme 31, 32,112, 210, 211, 290, 291,
296, 331, 332, 335, 336, 338, 357, 360
Isozyme marker 334
J
Japónica 12, 32, 36, 37,38, 40, 42, 43, 46,
107, 109,110, 111, 113,150,175,176,
180,188, 229, 230, 243, 251, 253, 276,
288, 291, 292, 293, 294, 296, 301, 331,
333, 335, 336,337353, 357, 358, 360, 361,
362,363, 364
Japonicoides 341
Javanica 32,33,112, 113,115, 331,333,
353,357,360,361,362', 364
K
Karotype 161,295
Karyomorphology 291
Keng 360,361
Kinetochores 302
Kresek 182,,183
Land race 353, 360
Laterite soil 289
latifoliae 314, 318,320, 325, 339,340, 341
Leaf area index 24
Leaf blast 170,171,178
Leaf folder 16,17,144,145,152,153,155,
211, 273, 289
Leaf hopper 152,189,246
Leaf scald 145
Leersia hexandra 209,321
Lethal genes 109
Likage disequilibrium 128,129,131,137
Linkage 30,41,128,129,150,174,175,
195, 241,247, 248, 280,282,293, 296,
299,300
Linkage mapping 296
Lodging 104
Lodging resistance 103,104,105,107
Lowland 2,3,10,14,46,64,65,70,73,74,
75,82,131,156,170, 223,230,233,361,
363, 364
M
Maintainer 35,36,38,39,42,46,113
Major gene effects 248
Major genes 177,180,181,248,261,262
Male sterility facilitated recurrent selection
(MSRS) 207,209
Mangrove 6,70,233
Mangrove swamp 220
Manure 57
Mapping 241, 242, 243, 246, 247,248, 249,
254, 261, 263, 282, 283

Ir\dex 377
Marker aided selection 160,203,210,241,
247,263, 264
Marker gene 296,299, .300
Markers 15,110,112,114,210,248,249,
251, 253, 262, 296,300,302,331,332
Mass selection 209
Mean 133
Mechanization 60
Mekong descent group 363
Mesophytes 314
Metacentric 298
Methane 7,12, 64, 86
Meyerianae 314, 322
Micronutrient 14
Micronutrient deficiency 5,13,56
Microprojectile bombardment 16
Microsporogenesis 38
Mid parent heterosis 24
Migration 357,362
Mining 57,58,70
Minor genes 181,185,187
Misdivision 297
Mitochondria 81
Mitochondrial DNA 336
Mixed cropping 7
Modified pedigree method 126
Modifier 171,174,179
Molecular genetic 287
Molecular hybridization 332
Molecular linkage 288, 299,302
Molecular marker 4,18,44,45,46,107,
160, 210, 211, 241, 242, 244,246, 247,
248, 249, 250, 251, 252, 263
Molecular probe 15
Molecular studies 331
Molecular tagging 282
Molecular techniques 181
Monocrop 62,71
Monoculture 56,60,145
Monogenic 42,109,146,147,175, 242, 243
Monograph 313
Monomorphic 280,331
Monophyletic origin 355
Monosomie alien addition line 276,288,
306, 307
Monosomies 301
Morphometric 334
Multi line 181
Multiple allele 170
Multiple cropping 70
Multiple nutrient stress 219
Multiple resistance 176,196, 206
Multivalent 330
Multivariate 31,32,132
Mutagenesis 250
Mutant 42,43,44,170
Mutation 114> 116,121, 209, 249,341
N
Narrow brown leaf spot 169
National Agriculture Research Stations
(NARS) 6,236
Natural enemies 58,145
Natural hybrid 351,353,354,355
Natural hybridization 316, 317, 351, 356
Natural resource 55
Natural selection 341
Near isogenic lines (NILs) 242,280
Neck blast 170,178, 179
Negative heterosis 24
Neutral allele 112
New plant type 105,106,107
Nitrifying bacteria 85
Nitrite oxide 12
Nitrogen fixation 57
Nodal root 88
Non allelic interaction 30
Non homologous 272,292
Non specific resistance 245
Nuclear 31
Nuclear DNA 331,334
Nuclear genome 336,337
Nuclear orgjmizing regions (NOR) 332
Nucleolus 292
Nucleolus organizer
Nutrient deficiency
Nutrient imbalance
Nutrient interaction
Nutrient management
Nutrients 58,101
O
292
89,220
224
225
57
O. abromeitiana 322
O. aha 272, 273, 289, 319, 320, 321, 323,
324, 325, 327, 328, 330, 331, 332, 335,
336.337.341.347
O. angustifolia 313,321,323,329,339,340,
347
O. australiensis 160,196,197,198,210,211,
246, 272, 273, 274, 279, 282, 289, 306,
307, 319, 320, 321, 323, 329, 331, 333,
334, 335, 337, 339, 347
O, barthii 36,188,209,316,317,318,323,
324.330.331.347

378 Rice Breeding and Genetics: Research Priorities and Challenges
O. brachyantha 160, 209, 210, 272, 273, 274,
279, 289, 314, 321, 323, 329, 331, 333,
338,339, 347
. breviligulata 272, 273,289,290,291,
316, 317, 324, 347
coarctata 313, 355
collina 320,325,326, 328
cubensis 350, 324, 347
eichengiri 209,272,273,289,319,320,
321, 323, 324, 325, 326, 327, 328, 332,
333, 335, 339, 340, 341, 347
fatua 351,355
glaberrima 36, 234,272, 273, 288,289,
290, 291, 294, 317,318, 315, 316,323,
324, 330,331, 341, 342, 347,349,350,355
glaberrima f. stapfii 316,324
glumaepatula 37, 272, 273, 275, 289,317,
318,323, 324, 330,333, 347350, 355
O . g 'ramlata 160,272,274, 288, 289, 290,
322, 331, 347
indandamanica 288,290,322
latifolia 160,190,209,210,272,273, 279,
289, 314,318, 319,320, 324, 325, 326,
327, 328, 330,331, 332, 333, 335, 336,
337, 339, 347
hngiglumis 272,274, 290,322,331,340,
347
longstaminata 36, 184,185,209,210,
211, 244, 272, 273, 274, 275, 280, 289,
291, 316, 317, 318, 323, 330, 331, 347,
350, 355,
malampzhuaensis 190, 319,320, 323,
324, 327, 328, 335, 340, 341
meridionalis 272, 273, 289,315,317,318,
323, 330, 335, 341, 347
meyeriam 209, 272,274, 288, 289, 290,
313, 314, 322, 323, 331, 335, 339, 340, 347
minuta 160, 178,190, 209, 210, 272, 273,
274, 278, 282, 289, 319, 320, 321, 323,
324, 326, 327, 328, 329, 331, 332, 335,
340, 341, 347354, 355,
nvara 36,37,181,192,193, 209, 210,
272, 273, 274, 275, 289, 290, 291, 315,
316, 317, 323, 330, 331, 335, 341, 347,
351, 352, 353, 354, 355, 356, 357, 358,
359, 360, 361, 362, 364
0 . officinalis 160,190, 209, 210, 272,273,
274, 277, 278, 280, 281, 282, 288, 289,
306, 307, 308, 318, 319, 320, 321, 323,
326, 327, 328, 330, 331, 332, 333, 334,
335, 337, 339, 340, 341, 347354, 355
o . paraguainensis 320, 324, 325, 329
O.perennis 36,37,272,273,274,275,317,
318, 324, 331, 350, 351, 355, 356
O. perennis subsp. balunga 317
O. perennis subsp. barlhii 317
O. perennis subsp. cubensis 317
O. perrieri 190,313, 314,321, 323,339,340,
347
O. punctata 209,272,273, 276,277, 288,
289,319, 320, 321, 323,326,327,331,
332, 333, 334, 341, 347
O, rhizomatis 272, 273,289, 318, 319,320,
327, 328, 340
O. ridleyi 209, 210, 272, 274, 288, 290, 314,
322,323,331,339, 340, 347
O. rufipogon 36,37,40,188,209,234, 272,
273, 275,289,290, 291,315,316,317,
318,323,324,330,331,332,333,335,
341,347,350,351,352,353, 354,355,
356,357,362
O. saliva 36,40,160,272, 273,276, 278,
279, 280, 281, 282,283,288, 289,290,
291, 294,306,307,308,314,315, 318,
323, 324,326,327,329,330,331, 332,
333, 334,335,336,341,342, 347, 349,
350,351, 353,354, 355,356, 361,362,
363, 364,
O, saliva var/fli«fl 351
O. schleeteri 272,288,290,314,322,323,
339,340, 347
O. schweinfurthiana 190,319,320,323,324,
326, 335, 341, 347
O.tisseranti 313,321, 323,339,340,347
Orange leaf 169
Organophosphorous compunds 145
Origin 31, 32,349,354,355,359,361
Oryza 272, 275, 287, 288, 290, 313, 314,
355, 321, 322, 323, 329, 332, 333, 334,
335,338,339
Oryza saliva f, spontanea 36,40,271,274,
275, 351, 315, 316, 324
Oryza species 122,196,274
Oryzeae 233, 288, 313
Oryzoideae 287, 288
Over dominance 28,30,136,253
Oxidative phosphorylation 89
Pachytene analysis 301
Pachytene chromosome 292,295,301,306
Pachytene idiogram 302
Pachytene trivalent 298

Index 379
Padia 314, 322,329,339, 340
Parallel variation 318, 350
Partial resistance 146,147,180,181,182,
243
Particle bombardment 16
Pathogen 4,146,1 5 3 ,1 5 4 ,1 7 0 ,1 8 4 , 229,
245
Pathogenicity 183,184,187,188,193
Pathotype 158,161,181,184,185
Peat soils 222,230,235
Pedigree 126,127,132
Pedigree method 121,125,209,229
Penetrance 135
Penyakit merah 199
Perennial 3 1 6 ,322,333,335,339,340,341
Perennial swamp 316
Perermial wild 352
Perennial wild species 350,355
Perermial species 318,352,356
P e r r ie r a n a e 314,321
Pest 36, 58, 6 9 ,1 2 0 ,122,143,144,145,153,
196,197, 246
Pesticide 4, 56,100,143,159, 209
Phenols 153,154
Phenotype 121,128,132,138,148, 299
Phenotypic 296
Phenotypic variation 262
Phnoloxidases 154
Phosphorous deficiency 222,225,226,
227, 229, 230, 232, 233, 236
Photoinsensitive 145,362
Photoperiod 2,249
Photoperiod insensitive 315,353,359,361
Photoperiod sensitive 82,229,234,261,
262, 352, 353, 362
Photoperiod sensitive genetic male sterility
(PGMS) 3 ,4 2 ,4 3 ,2 4 9 ,2 5 0
Photophosphorylation 79
Photorespiration 78,79
Photosensitive 175,231,315,358
Photosynthate 103
Photosynthesis 7 7 ,7 8 ,7 9 ,8 1 ,8 4 ,8 5 ,9 0 ,
101,104,106
Phylogenetic 290,333,339,357
Phylogenetic relationships 331,334,338
Physical map 293,334
Phytoalexine 154
Plant genome 160
Plant hopper 152,153,246
Plant ideotype 11
Plant nutrients 54, 57, 234
Plant type 2 ,1 1 ,1 2 ,5 9 ,6 8 ,1 0 0 ,1 0 1 ,1 0 5 ,
1 2 0 ,1 2 1 ,1 24,131,190,191,192,205, 207
Planthopper 144
Pleiotropy 128,129,210
Ploidy 330
Poaceae 287,288
Polyacrylamide gel electrophoresis
(PAGE) 331
Polyethylene glycol (PEG) 15,16
Polygenes 5 9 ,1 5 1 ,162,172,173,175,177,
■ 180,185,205,211
Polygenic resistance 147
Polymorphism 280,291,335,337
Polyphagous pest 147
Polyphyletic origin 354,363
Polyploid 335
Polyploidization 335
Poly topic 363
Population 1 ,2 ,9 ,1 0 ,5 3 ,6 5 ,7 1 ,9 9 ,1 1 9 ,
128,138,161,162,171,206,208, 219,
261,334, 357,359,361
Porleresia coarctata 209,233,355
Primary gene pool 209,272,274
Primary trisomics 288,293,294,295, 296,
2 9 7 ,298,299,301,306,307
Primitive 360
Primitive cultivar 362
Primitive ecotype 358
Primitive species 338
Problem soils 54,219,220,231,235
Progenitor 357,361
Projectile bombardment 211
Prophylaxis 59
Protease 15,17
Protoclonal variation 15
Protoplast 15
Protoplast fusion 15,16,44, 45
Pure line selection 209
Putative progenitors 354
Pyramiding 147,160,180,181,189,263
Pyrethroids 145
Q
Quadrivalent 322, 329
Qualitative 146,147,151, 243
Qualitative resistance 245
Quantitative 2 8 ,3 0 ,1 7 3 ,2 1 1 ,2 4 6 , 247
Quantitative characters 122,125,132, 254
Quantitative resistance 146,156,243,245
Quantitativetraitloci(QTL) 149,181, 211,
243, 245, 246, 247, 250, 252, 253, 254,
261, 262, 264, 265, 283

380 Rice Breeding and Genetics: Research Priorities and Challenges
R
Race non specific resistance 146,147
Races 150,158,176,188, 272, 273, 279,307
Race specific resistance 146,147
Ragged stunt 169
Rainfed 2,10,13,14,46, 64, 65,70,73, 74,
75,80,82, 90,144,156,222,359,361
Rainfed systems 10
Random amplified polymorphic DNA
(RAPD) 32, 242, 251, 280, 282
Random chromatid segregation 296
Random chromosome segregation 296
Rapid generation advance (RGA) 231
Rate reducing resistance 146
Ratooning 25,353
Recessive 42, 135,148,149,150,151,174,
185,187,194,195,199,200,202,210,
251, 299,300
Recessive effects 177
Reciprocal effects 233
Reciprocal translocation 301
Recurrent selection 206
Repetitive DNA 161,296,337, 338
Repetitive sequences for genome
specificity 331,337
Resistance 59, 83,101,135,143,147,149,
150,152, 153,154,156,158,160,169,
170,171,172,174,175,177,178,185,
187,190,191, 195,196,197,198,199,
200,201, 205, 206,207,208,227,229,
242, 243, 264, 273, 274, 277, 279, 280,
282, 289
Restitution nuclei 326
Restorer 37,38,39,40, 41,42,46,113
Restriction analysis 331,334
Restriction Endonuclease Analysis 333
Restriction fragment length polymorphism
(RFLP) 31,161,162, 210, 211, 242, 248,
251, 280, 281, 290, 296, 299, 301, 306,
334,335,336, 338
Rhizomatous 289
Rhizosphere 85
Ribosomal DNA (rDNA) 293,296,332,
333, 338
Ribosomal gene 338
Ribosome activity enzyme (RIP) 16
Rice 1, 2, 5,10, 25, 26, 27, 30, 34, 39,42,53,
54, 56, 62, 99,104, 271, 275
Rice based cropping 66
Rice cropping system 63,145
Rice dwarf 199, 200
Rice farming 60,61, 62, 64,68, 71
Rice gall dwarf 200
Rice production '73
Ridleyanae 314, 322, 340
Rodents 58
Root respiration 81,82,85
Root system 104
Saline 219,232,235
Saline soils 221
Salinity 13,56,60,63,220,221,225,226,
227,229, 231,233, 236,272
Salinity tolerant 355
Salinization 4
Salt injury, 227
Salt tolerance 14,25, 230, 233
Satellite chromosome 332
Sativa 315
Sativae 314, 317, 318, 339, 341
Schtechierianae 314,322
Screening 225,236
Secondary chromosome 297
Secondary ecotypes 361,362
Secondary gene pool 272,275
Secondary trisomics 288,297,298,299,
300
Seed 70
Seihr 353,357,358,359, 360,361,362,363
Selection 125,128
Selection efficiency 210
Selection index 138
Selection pressure 121,354
Semi-dwarf 1,3,12, 45, 59, 68, 99,100,
229,231, 261
Semi-sterile 110,114
Shade tolerance 274,289
Shall 353,358,362,363
Shallow rainfed lowland 14
Sheath blight, (ShB) 16,17,105,144,158,
161,169,187,189, 245, 271, 273, 274, 289
Sheath rot 144,145
Shifting cultivation 55, 359
Shuttle breeding 231
Silicon deficiency 223,232,235
Silver shoot 202
Simple sequence repeats (SSRs) 242,290
Single seed descent (SSD) 123,124,125,
126,127,132,133,135, 209
Sínica 360
Sink 79,80,101,102,106
Slash and burn 55
Slot-blot-hybridization 337

Index 381
Socioeconomic 54, 63,67, 209
Sodic 219,235
Sodicity 221,228
Soil 104,235,274
Soil acidity 14,183
Soil amendments 234,235
Soil erosion 14, 66
Soil fertility 57,75
Soil nitrogen 63
Solar energy 103
Solar radiation 104
Somatic embryogenesis 44,45
Somoclonal variation 15,229
Southern hybridization 282,333
Specific resistance 146,150
Spontanea 354
Sporophyte 35,37,41
Spring rice 62
Sress tolerance 262,263
Stability 36,135,136
Stabilizing selection 171
Stable habitat 352
Stalked eye fly 204
Standard evaluation system (SES) 206,
226
Standard heterosis 24,29
Starch biosynthesis 18
Stem borer 16,17,105,144,145,147,148,
152,153,157,160, 203, 205, 206,207,
208, 211, 272, 273, 274, 279, 280, 289, 290
Stem elongation 87
Sterility 33,35
Stoloniferous 290
Stripe virus 244
Striped borer 206, 208
Subgenomic differentiation 330
Submerged soil 224
Submergence 13,14, 73, 74, 84, 85, 88,89,
90, 272
Submersion 62,64
Submetacentric 298
Sulfur deficiency 223
Super rice 102,105
Sustainability 54, 55,56,61,63,64,65,66,
68,69
Swamp 352
Sympatric 354,362,340
Synapsis 330
Tagging 236,244, 246,247,248,263
Tagging gene 210
Taxa 317,324,350
Taxon 351
Taxonomic 317,350, 353, 354, 356
Taxonomy 287,314,331
Telo trisomics 288, 297, 298, 299
Telocentric 297, 299
Temperate japónica 32,107
Temperature sensitive genetic male
sterility (TGMS) 3,42,43,250,251
Test cross 115
Tester 112,174,324
Tetraploid 44,282,288,290,355,314,318,
319,320,322,323,324,325,332,335,340
Tetrasomics 301
Thionins 17
Thrips 273
Thriving with rice 5
Tidal swamp 2,355
Tidal wetland 14,64
Tillage 58
Tissue culture 14,283
Tjereh 41,353,358,362,363
Tolerance 227,229,233
Toxicity 219,221,231, 234
Trans gene silencing 264
Transformation 15,44,45,161,263,280,
283
Transgenic 4,15,16,17,160,211,264,280
Transgressive segregants 122,124,127,
'131,137
Transgressive segregation 233,245
Transgressive variation 124,132
Transitory yellowing 200
Translocation 41,79,103
Translocations 288
Transmission 295,299,306,307
Transpiration 76,80,81
Transposable elements 162
Trap crop 59
Triple test cross 133
Triplo 294,295,297
Triploid 44
Trisomie 41,195,248,332
Trivalent 292, 298, 307, 322, 327
Tropical japónica 12,46,107,120,232
True resistance 180
Tungro 18,59,105,107,144,151,157,158,
161,169,189,190,191,192,199,201,
244, 246
Two line rice hybrids 43,113

382 Rice Breeding and Genetics: Research Priorities and Challenges
U
Pnited Nations Conference on
Environment and Development
(UNCED) 7
Univalent 297, 298, 322, 327, 329
Upland 2, 7,10,11,13, 65, 66, 67, 70, 73,
74, 76, 81, 82,170, 202, 222, 223, 231,
359, 361
Uruguay Round of Multilateral Trade
Negotiation 7
Variance 133,134
Variance analysis 123, 252, 262
Variance components 123
Vector 59, 190,274
Vertical resistance 146
Vertifolia effect 181
Virulence 171,187, 243
Virulent 153,170
Virus 58, 59,105,107,157,161,190,193,
195, 200, 242, 246, 274
W
Water deficit 76, 77
Water management 56,58, 65
Water use efficiency 57,61,68,80,81
Waterlogging 56,60,86,229
Weeds 4,14,58,59,63,64,65, 74,120
West African Rice Development
Association (WARDA) 224,232,236
Wetland 86,223,362
Wheat 5, Ip
White back plant hopper (WBPH) 144,
148,149,157,160,197,198,199, 211,
271, 273, 274, 277, 278, 279, 289, 290
White borer 204
White heads 204
Whorl maggot 273,279,289,290
Wide compatibility 336
Wide compatibility gene (WCG) 33,46,
113, 251
Wide compatibility type 111, 250,251
Wide compatibility variety (WCV) 34,
109,112, 251,336
Wide crosses 44
Wide hybridisation 14,122,209
Wild abortive (WA) 248,275, 275,248,31,
36,39,41
Wild rices 316
Wild species 354,315
Yellow dwarf 199,200
Yellow mottle virus 273,289
Yellow orange leaf 199,200,201
Yield 3, 27,31,33,77,101,103,104,121,
125, 254,264,265
Yield components 25,29,254,264,265
Yield potential 1,53,60,99,100,102,103,
106,122,131, 278
Zigzag leafhopper (ZLH) 197,198,201,
273,274
Zinc deficiency 222,225,226,227,228,
229,230,231,235,236